@mengyuly/dsh-ponytail 0.3.1 → 0.3.3

package/lib/index.js

@@ -1,1795 +1,1748 @@
-import { createRequire } from "node:module";
-import { Service } from "@deepseek-ai/cordis";
-import z from "@deepseek-ai/schemastery";
-import { mkdirSync, readFileSync, renameSync, unlinkSync, unwatchFile, watchFile, writeFileSync } from "node:fs";
-import { homedir } from "node:os";
-import { dirname, join } from "node:path";
-//#region ../../llm/llm/src/brand.ts
-/**
-* Brand a message identifier.
-* @param id - the opaque message identifier.
-* @returns the same string, branded; no validation is performed.
-*/
-function MessageId(id) {
-	return id;
-}
-//#endregion
-//#region ../../llm/llm/src/call-config.ts
-/**
-* Deep-freeze a value in place with an iterative traversal, guarding cycles,
-* so later mutation throws without imposing a JavaScript call-stack depth cap.
-* {@link AbortSignal} objects are deliberately skipped because they are the
-* request's live cancellation channel and freezing them breaks abort.
-* @param value - the value to freeze in place.
-* @returns the same value, frozen.
-*/
-function deepFreeze(value) {
-	const seen = /* @__PURE__ */ new WeakSet();
-	const pending = [{
-		kind: "visit",
-		node: value
-	}];
-	while (pending.length > 0) {
-		const task = pending.pop();
-		/* v8 ignore next -- the loop condition guarantees one pending task. */
-		if (task === void 0) continue;
-		if (task.kind === "property") {
-			pending.push({
-				kind: "visit",
-				node: task.source[task.key]
-			});
-			continue;
-		}
-		const node = task.node;
-		if (node === null || typeof node !== "object") continue;
-		if (node instanceof AbortSignal) continue;
-		if (seen.has(node)) continue;
-		seen.add(node);
-		Object.freeze(node);
-		const keys = Object.keys(node);
-		for (let index = keys.length - 1; index >= 0; index--) {
-			const key = keys[index];
-			/* v8 ignore next -- the loop is bounded by the captured key count. */
-			if (key === void 0) continue;
-			pending.push({
-				kind: "property",
-				source: node,
-				key
-			});
-		}
-	}
-	return value;
-}
-//#endregion
-//#region ../../llm/llm/src/message.ts
-/** Message value types, identity, and immutable construction helpers. */
-/**
-* Detach and deep-freeze a message whose identity already exists.
-* @param message - complete message, including its stable identity.
-* @returns an immutable snapshot that preserves the identity.
-*/
-function freezeMessage(message) {
-	return deepFreeze(structuredClone(message));
-}
-/**
-* Create one identified message and freeze it before publication.
-* @param input - complete role, content, and source for a new message.
-* @returns an immutable message with a fresh stable identity.
-*/
-function createMessage(input) {
-	return freezeMessage({
-		...input,
-		id: MessageId(crypto.randomUUID())
-	});
-}
-/**
-* Create one identified user-role message and freeze it before publication.
-* @param input - complete content and source for a new user message.
-* @returns an immutable user message with a fresh stable identity.
-*/
-function createUserMessage(input) {
-	return createMessage({
-		...input,
-		role: "user"
-	});
-}
-//#endregion
-//#region ../../util/timeout/src/index.ts
-/** Largest delay Node schedules without clamping it to one millisecond. */
-const MAX_TIMER_DELAY_MS = 2147483647;
-//#endregion
-//#region ../../llm/llm/src/error.ts
-/**
-* Canonical provider-neutral code for a response that completed normally but
-* carried no content blocks at all. Providers occasionally emit a degenerate
-* completion (a terminal stop with zero output); adapters classify it as this
-* failure instead of yielding an empty assistant message, because an empty
-* message silently ends the turn with nothing for the user or the loop to act
-* on. The attempt produced nothing durable, so retry policy treats it as safe
-* to repeat.
-*/
-const EMPTY_RESPONSE_CODE = "EMPTY_RESPONSE";
-new RegExp(String.raw`(?:^|[^a-z0-9])context[\s_-](?:length|window)[\s_-]` + String.raw`(?:exceed(?:ed|s)?|overflow(?:ed)?|limit[\s_-]exceeded)(?:$|[^a-z0-9])`, "i");
-new RegExp(String.raw`\b(?:request|prompt|input|messages?)\s+(?:is\s+|are\s+)?` + String.raw`too\s+(?:large|long)\s+for\s+(?:(?:this|the)\s+)?` + String.raw`(?:model(?:'s)?\s+)?context(?:\s+window)?\b`, "i");
-new RegExp(String.raw`\b(?:input|prompt|request|messages?)\b.{0,40}` + String.raw`\b(?:exceed(?:s|ed)?|overflows?|is\s+larger\s+than)\b.{0,40}` + String.raw`\b(?:the\s+)?(?:model(?:'s)?\s+)?context(?:\s+(?:length|window))?\b`, "i");
-//#endregion
-//#region ../../llm/llm/src/retry-policy.ts
-/**
-* Provider-owned request-retry policy configuration and resolution.
-*
-* Adapters expose one resolved policy per registered provider route; the
-* optional dsh-llm-retry plugin executes it on the agent's failed-step extension point.
-*
-* @module @deepseek-ai/dsh-llm/retry-policy
-*/
-const DEFAULT_MAX_RETRIES = 5;
-const DEFAULT_INITIAL_DELAY_MS = 500;
-const DEFAULT_MAX_DELAY_MS = 1e4;
-const DEFAULT_JITTER_RATIO = .1;
-const DEFAULT_RETRYABLE_CODES = Object.freeze([
-	EMPTY_RESPONSE_CODE,
-	"RATE_LIMIT",
-	"SERVER",
-	"TIMEOUT",
-	"TRANSPORT"
-]);
-const backoffSchema = z.object({
-	initialDelayMs: z.number().max(MAX_TIMER_DELAY_MS).default(DEFAULT_INITIAL_DELAY_MS),
-	maxDelayMs: z.number().max(MAX_TIMER_DELAY_MS).default(DEFAULT_MAX_DELAY_MS),
-	jitterRatio: z.number().min(0).max(1).default(DEFAULT_JITTER_RATIO)
-});
-const normalPolicySchema = z.object({
-	mode: z.const("normal").required(),
-	maxRetries: z.number().step(1).min(0).max(Number.MAX_SAFE_INTEGER).default(DEFAULT_MAX_RETRIES),
-	retryableCodes: z.array(z.string()).default([...DEFAULT_RETRYABLE_CODES]),
-	backoff: backoffSchema
-});
-const alwaysPolicySchema = z.object({
-	mode: z.const("always").required(),
-	backoff: backoffSchema
-});
-z.union([normalPolicySchema, alwaysPolicySchema]);
-//#endregion
-//#region ../../llm/llm/src/attribution.ts
-/**
-* Centralize the non-secret product identity every provider request sends as `User-Agent`, keeping
-* adapters from drifting. See
-* `.agents/notes/implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md`.
-*
-* App-attribution vocabulary for provider requests.
-* @module @deepseek-ai/dsh-llm/attribution
-*/
-const { version } = createRequire(import.meta.url)("../package.json");
-//#endregion
-//#region ../../llm/llm/src/never.ts
-/**
-* Exhaustiveness helper for closed core unions. Use {@link assertNever} at the default branch so a
-* new variant fails compilation at every required handler. Do not use it for declaration-merged
-* unions such as session events or content blocks: handle known variants and explicitly fall
-* through because plugins may add valid unknown cases.
-* @module @deepseek-ai/dsh-llm/never
-*/
-/**
-* Mark an unreachable closed-union branch. A newly unhandled typed variant fails at the call site;
-* a value that escaped its type throws with diagnostics at runtime.
-* @param value - the impossible value; typed `never` so an unhandled variant fails compilation at the call site.
-* @param context - optional label (e.g. the switch site) prefixed into the throw message.
-* @returns never — it always throws, with the offending value JSON-rendered in the message.
-*/
-function assertNever(value, context) {
-	const rendered = JSON.stringify(value) ?? String(value);
-	throw new Error(`unreachable variant${context ? ` in ${context}` : ""}: ${rendered}`);
-}
-//#endregion
-//#region ../../core/scope/src/store.ts
-/**
-* Insertion-ordered named entries with caller-owned duplicate diagnostics.
-*
-* Values are borrowed. Iterators are live within one nonempty table
-* generation; draining the table detaches them from later insertions. Each
-* successful insertion returns an idempotent undo for that exact entry.
-*/
-var NamedEntries = class {
-	duplicateError;
-	data = /* @__PURE__ */ new Map();
-	constructor(duplicateError) {
-		this.duplicateError = duplicateError;
-	}
-	/**
-	* Insert one unique name.
-	* @param name - name unique within this table.
-	* @param value - borrowed value to retain.
-	* @returns an idempotent undo that removes only this insertion.
-	*/
-	insert(name, value) {
-		const data = this.data;
-		if (data.has(name)) throw this.duplicateError(name);
-		data.set(name, value);
-		let active = true;
-		return () => {
-			if (!active) return;
-			active = false;
-			data.delete(name);
-			if (data.size === 0 && this.data === data) this.data = /* @__PURE__ */ new Map();
-		};
-	}
-	/**
-	* Read one named value.
-	* @param name - name to resolve.
-	* @returns the retained value, or `undefined` when absent.
-	*/
-	get(name) {
-		return this.data.get(name);
-	}
-	/**
-	* Test one name for membership.
-	* @param name - name to test.
-	* @returns whether the table contains that name.
-	*/
-	has(name) {
-		return this.data.has(name);
-	}
-	/**
-	* Iterate live names in insertion order.
-	* @returns the native live key iterator.
-	*/
-	keys() {
-		return this.data.keys();
-	}
-	/**
-	* Iterate live entries in insertion order.
-	* @returns the native live entry iterator.
-	*/
-	entries() {
-		return this.data.entries();
-	}
-	/**
-	* Iterate live values in insertion order.
-	* @returns the native live value iterator.
-	*/
-	values() {
-		return this.data.values();
-	}
-	/**
-	* Test whether this table has no entries.
-	* @returns whether the table is empty.
-	*/
-	isEmpty() {
-		return this.data.size === 0;
-	}
-};
-/**
-* Own the global and exact-scope layers for one registry.
-*
-* Reads never create scoped layers. Registrations derive both visibility and
-* effect ownership from the supplied Cordis context, collect undo before
-* notification, and reclaim only a completely empty aggregate layer.
-*/
-var ScopedLayers = class {
-	createLayer;
-	onChange;
-	/** The eagerly constructed context-global layer. */
-	global;
-	scoped = /* @__PURE__ */ new Map();
-	constructor(createLayer, onChange) {
-		this.createLayer = createLayer;
-		this.onChange = onChange;
-		this.global = createLayer(void 0);
-	}
-	/**
-	* Read an existing exact-scope overlay. Deliberately chain-blind: callers
-	* addressing one scope's OWN contributions (its restrictions, its guards)
-	* must not silently pick up an ancestor's — use {@link chainLayers} where
-	* inheritance is the point.
-	* @param scope - exact scope key; `undefined` denotes no overlay.
-	* @returns the existing scoped layer, or `undefined` without creating one.
-	*/
-	peek(scope) {
-		if (scope === void 0) return void 0;
-		return this.scoped.get(scope);
-	}
-	/**
-	* Existing overlays along the scope's parent chain ({@link scopeChainOf}),
-	* farthest ancestor first and the exact scope last, so a caller layering
-	* them in order gives the nearest scope the final word.
-	* @param scope - viewing scope, or `undefined` for no overlays.
-	* @returns the existing layers, nearest last; absent overlays are skipped.
-	*/
-	chainLayers(scope) {
-		const layers = [];
-		for (const key of scopeChainOf(scope).reverse()) {
-			const layer = this.scoped.get(key);
-			if (layer !== void 0) layers.push(layer);
-		}
-		return layers;
-	}
-	/**
-	* Materialize global named entries followed by scope-chain shadows,
-	* farthest ancestor first, so the nearest scope's entry wins a name.
-	* @param scope - viewing scope, or `undefined` for the global view.
-	* @param pick - select the named table from a layer.
-	* @returns an insertion-ordered effective map.
-	*/
-	merge(scope, pick) {
-		const merged = new Map(pick(this.global).entries());
-		for (const layer of this.chainLayers(scope)) for (const [name, value] of pick(layer).entries()) merged.set(name, value);
-		return merged;
-	}
-	/**
-	* Attach one synchronous layer mutation to its registration context.
-	* @param ctx - context that determines both scope visibility and effect ownership.
-	* @param action - atomic mutation returning its synchronous undo.
-	* @param options - Cordis effect label and optional change notification.
-	* @returns the exact disposer returned by `ctx.effect()`.
-	*/
-	effect(ctx, action, options) {
-		const scope = scopeOf(ctx);
-		const notify = options.notify ?? true;
-		return ctx.effect(function* () {
-			let layer;
-			let created = false;
-			if (scope === void 0) layer = this.global;
-			else {
-				const existing = this.scoped.get(scope);
-				if (existing === void 0) {
-					layer = this.createLayer(scope);
-					this.scoped.set(scope, layer);
-					created = true;
-				} else layer = existing;
-			}
-			let undo;
-			try {
-				undo = action(layer);
-			} catch (error) {
-				if (scope !== void 0 && created && layer.isEmpty()) this.scoped.delete(scope);
-				throw error;
-			}
-			yield () => {
-				undo();
-				if (scope !== void 0 && layer.isEmpty()) this.scoped.delete(scope);
-				if (notify) this.onChange();
-			};
-			if (notify) this.onChange();
-		}.bind(this), options.label);
-	}
-};
-//#endregion
-//#region ../../core/scope/src/index.ts
-/** Context tag written by {@link createScope}. */
-const kScope = Symbol("dsh.scope");
-/**
-* The enclosing scope of each key. One relation powers both directions of
-* scope nesting: registration views inherit DOWN the chain (a child scope
-* sees its ancestors' layers — {@link ScopedLayers}), and event admission
-* extends UP it (a listener tagged with an ancestor receives events dispatched
-* to a descendant key — {@link scopeTarget}).
-*/
-const scopeParents = /* @__PURE__ */ new WeakMap();
-/**
-* The chain from a key to its root ancestor.
-* @param key - the starting key, or `undefined` for the empty chain.
-* @returns keys nearest-first: `[key, parent, grandparent, …]`.
-*/
-function scopeChainOf(key) {
-	const chain = [];
-	for (let cursor = key; cursor !== void 0; cursor = scopeParents.get(cursor)) chain.push(cursor);
-	return chain;
-}
-/**
-* Read the nearest scope tag inherited by a context.
-* @param ctx - context to inspect.
-* @returns its scope key, or `undefined` for an unscoped context.
-*/
-function scopeOf(ctx) {
-	return ctx[kScope];
-}
-//#endregion
-//#region ../../skill/skill/src/index.ts
-/**
-* Agent skill provider registry.
-*
-* This package owns the Service Definition role of the skill capability seam.
-* Concrete
-* providers such as `@deepseek-ai/dsh-skill-filesystem` decide where skills come
-* from; this service only merges provider catalogs, resolves the winning skill
-* for a name, and exposes the winning summaries and definitions to consumers.
-*
-* @module @deepseek-ai/dsh-skill
-*/
-const SKILL_NAME = /^[a-z0-9]+(?:-[a-z0-9]+)*$/;
-const DEFAULT_COLLECT_CACHE_ENTRIES = 128;
-const MAX_COLLECT_ATTEMPTS = 2;
-const RUNTIME_PROVIDER = "runtime";
-const RUNTIME_RANK = 250;
-/**
-* Return whether a string is a valid kebab-case skill name.
-* @param name - candidate skill name to validate.
-* @returns whether the name matches the public skill-name grammar.
-*/
-function isSkillName(name) {
-	return SKILL_NAME.test(name);
-}
-/**
-* Render one loaded skill for the model. The output is shared verbatim by the
-* `skill` tool result and the user-explicit invocation injection, so the model
-* sees one canonical `<skill_content>` shape on both paths. The name rides an
-* escaped attribute; the body is embedded verbatim (skills are trusted local
-* content, and user-supplied invocation text stays outside this wrapper).
-* @param skill - name, provider, optional resource base, and body to render.
-* @returns the complete model-facing `<skill_content>` block.
-*/
-function renderSkillContent(skill) {
-	const resourceHint = renderResourceHint(skill);
-	return [
-		`<skill_content name="${escapeAttr(skill.name)}">`,
-		"<skill_resources>",
-		...resourceHint,
-		"</skill_resources>",
-		"",
-		"<skill_instructions>",
-		skill.content,
-		"</skill_instructions>",
-		"</skill_content>"
-	].join("\n");
-}
-function renderResourceHint(skill) {
-	const base = skill.resourceBase;
-	if (base === void 0) return [`Resources for this skill are managed by provider "${escapeText(skill.provider)}".`, "Load referenced resources only as needed."];
-	switch (base.kind) {
-		case "directory": return [`Base directory for this skill: ${escapeText(base.path)}`, "Resolve relative paths mentioned by this skill against the base directory before using them. Load referenced resources only as needed."];
-		case "url": return [`Base URL for this skill: ${escapeText(base.url)}`, "Resolve relative URLs mentioned by this skill against the base URL before using them. Load referenced resources only as needed."];
-		case "opaque": return [`Resources for this skill: ${escapeText(base.description)}`, "Load referenced resources only as needed."];
-		/* v8 ignore start -- SkillResourceBase is a closed union; a future kind must fail compilation here. */
-		default: return assertNever(base, "SkillResourceBase.kind");
-	}
-}
-function escapeAttr(value) {
-	return value.replaceAll("&", "&amp;").replaceAll("\"", "&quot;").replaceAll("<", "&lt;");
-}
-/**
-* Escape model-facing prose embedded inside skill markup so provider-supplied
-* text cannot open or close framing tags.
-* @param value - raw prose to embed.
-* @returns the escaped text.
-*/
-function escapeText(value) {
-	return value.replaceAll("&", "&amp;").replaceAll("<", "&lt;").replaceAll(">", "&gt;");
-}
-/** One scope's complete skill-registry contribution. */
-var SkillLayer = class {
-	/** Providers registered through contexts carrying this scope, insertion-ordered. */
-	providers;
-	/** Runtime skills registered through contexts carrying this scope. */
-	runtime = /* @__PURE__ */ new Map();
-	constructor(scope) {
-		this.providers = new NamedEntries((name) => /* @__PURE__ */ new Error(scope === void 0 ? `a skill provider named "${name}" is already registered` : `a skill provider named "${name}" is already registered in this scope`));
-	}
-	/** Whether every contribution table in this aggregate layer is empty. */
-	isEmpty() {
-		return this.providers.isEmpty() && this.runtime.size === 0;
-	}
-};
-(class extends Service {
-	static Config = z.object({ collectCacheMaxEntries: z.number().default(DEFAULT_COLLECT_CACHE_ENTRIES) });
-	collectCacheMaxEntries;
-	layers = new ScopedLayers((scope) => new SkillLayer(scope), () => {
-		this.invalidateCache();
-	});
-	collectCache = /* @__PURE__ */ new Map();
-	revision = 0;
-	nextProviderOrder = 0;
-	/** Stable identities for cache keys; scope keys are opaque identity-compared objects. */
-	scopeIds = /* @__PURE__ */ new WeakMap();
-	nextScopeId = 1;
-	constructor(ctx, config = {}) {
-		super(ctx, "skills");
-		this.collectCacheMaxEntries = config.collectCacheMaxEntries ?? DEFAULT_COLLECT_CACHE_ENTRIES;
-		assertPositiveInteger("collectCacheMaxEntries", this.collectCacheMaxEntries);
-	}
-	/**
-	* Register a borrowed same-process provider synchronously during plugin
-	* apply, into the calling context's layer: a scoped context (an agent
-	* preset's standing mount) registers for that scope alone, an unscoped
-	* context registers globally. Duplicate names within one layer and reserved
-	* names throw; remote initialization belongs in `list()`. Fiber disposal
-	* unregisters the provider and invalidates catalog caches.
-	* @param create - synchronous factory receiving this registration's lifecycle and invalidation control.
-	* @returns the exact Cordis effect disposer that unregisters this provider;
-	*   composite effects may yield it directly to preserve teardown ordering.
-	*/
-	registerProvider(create) {
-		const lifecycle = new AbortController();
-		let registration;
-		let provider;
-		const control = {
-			signal: lifecycle.signal,
-			invalidate: () => {
-				const active = registration;
-				if (active !== void 0 && active.layer.providers.get(active.name)?.provider === provider) this.invalidateCache();
-			}
-		};
-		try {
-			provider = create(control);
-			const name = provider.name;
-			if (name === RUNTIME_PROVIDER) throw new Error(`"${RUNTIME_PROVIDER}" is reserved for runtime skill registrations`);
-			const order = this.nextProviderOrder;
-			this.nextProviderOrder += 1;
-			return this.layers.effect(this.ctx, (layer) => {
-				const undo = layer.providers.insert(name, {
-					provider,
-					order
-				});
-				registration = {
-					layer,
-					name
-				};
-				return () => {
-					registration = void 0;
-					undo();
-					lifecycle.abort(/* @__PURE__ */ new Error(`skill provider "${name}" disposed`));
-				};
-			}, { label: "skills.registerProvider()" });
-		} catch (error) {
-			lifecycle.abort(error);
-			throw error;
-		}
-	}
-	/**
-	* Register a borrowed readonly runtime skill into the calling context's
-	* layer. Project entries outrank runtime entries, which outrank user
-	* entries, within one layer. Same-name runtime entries in one layer are
-	* first-wins; a duplicate logs a warning and receives a no-op disposer so
-	* it cannot remove the winner.
-	* @param skill - the skill definition input; omitted invocation and provider fields receive defaults.
-	* @returns the exact Cordis effect disposer, preserving composite teardown order and invalidating caches.
-	*/
-	register(skill) {
-		validateRuntimeSkill(skill);
-		const scope = scopeOf(this.ctx);
-		const existingLayer = scope === void 0 ? this.layers.global : this.layers.peek(scope);
-		if (existingLayer !== void 0 && existingLayer.runtime.has(skill.name)) {
-			this.ctx.logger.warn(`runtime skill "${skill.name}" ignored because it is already registered`);
-			return () => {};
-		}
-		const definition = {
-			...skill,
-			invocation: skill.invocation ?? {
-				modelInvocable: true,
-				userInvocable: true
-			},
-			provider: skill.provider ?? RUNTIME_PROVIDER
-		};
-		return this.layers.effect(this.ctx, (layer) => {
-			layer.runtime.set(definition.name, definition);
-			return () => {
-				layer.runtime.delete(definition.name);
-			};
-		}, { label: "skills.register()" });
-	}
-	/**
-	* List invocation-neutral skill summaries for a workspace. Consumers apply
-	* model or user invocation policy at their operational boundary. Lookup
-	* options and provider candidates are readonly same-process values borrowed
-	* throughout discovery.
-	* @param options - view options; `scope` selects the viewing agent's layers, `cwd` selects project roots, and `signal` cancels discovery.
-	* @returns all sorted winning summaries.
-	*/
-	async list(options = {}) {
-		return (await this.snapshot(options)).skills;
-	}
-	/**
-	* Observe the current invocation-neutral catalog and whether discovery completed within a stable revision.
-	* Incomplete observations are never cached, allowing consumers to retain last-good state and
-	* retry on their next request boundary.
-	* @param options - view options; `scope` selects the viewing agent's layers, `cwd` selects project roots, and `signal` cancels discovery.
-	* @returns sorted summaries plus discovery-completeness state.
-	*/
-	async snapshot(options = {}) {
-		const collected = await this.collect(options);
-		return {
-			skills: [...collected.entries.values()].map((entry) => toSummary(entry.candidate)).sort(compareSkillSummary),
-			complete: collected.cacheable
-		};
-	}
-	/**
-	* Load and validate the winning candidate, passing its opaque discovery locator back to the
-	* provider. Cancellation is rechecked after selection, including cache hits, and raced against
-	* loading so an uncooperative provider cannot hang the caller.
-	* @param name - kebab-case skill name.
-	* @param options - view options; `scope` selects the viewing agent's layers,
-	*   `cwd` selects workspace-sensitive skills, and `signal` cancels work.
-	* @returns the full skill, including body content, or `undefined`.
-	*/
-	async get(name, options = {}) {
-		if (!isSkillName(name)) return void 0;
-		const collected = await this.collect(options);
-		throwIfAborted(options.signal);
-		const match = collected.entries.get(name);
-		if (match === void 0) return void 0;
-		const definition = await waitWithAbort(match.provider.get(match.candidate, options), options.signal);
-		if (definition === void 0) return void 0;
-		validateDefinition(definition);
-		if (definition.name !== match.candidate.name) {
-			this.invalidateEntry(match);
-			return;
-		}
-		return definition;
-	}
-	async collect(options) {
-		throwIfAborted(options.signal);
-		let attempt = 1;
-		while (true) {
-			const revision = this.revision;
-			const key = this.collectCacheKey(options.cwd, scopeChainOf(options.scope), revision);
-			const cached = this.collectCache.get(key);
-			if (cached !== void 0) return {
-				entries: cached,
-				cacheable: true
-			};
-			const result = await this.collectFresh(options);
-			throwIfAborted(options.signal);
-			if (revision !== this.revision) {
-				if (attempt < MAX_COLLECT_ATTEMPTS) {
-					attempt += 1;
-					continue;
-				}
-				return {
-					entries: result.entries,
-					cacheable: false
-				};
-			}
-			if (result.cacheable) {
-				this.collectCache.set(key, result.entries);
-				if (this.collectCache.size > this.collectCacheMaxEntries) {
-					const oldest = this.collectCache.keys().next();
-					this.collectCache.delete(oldest.value);
-				}
-			}
-			return result;
-		}
-	}
-	async collectFresh(options) {
-		const layers = [this.layers.global, ...this.layers.chainLayers(options.scope)];
-		const merged = /* @__PURE__ */ new Map();
-		let cacheable = true;
-		for (const layer of layers) {
-			const collected = await this.collectLayer(layer, options);
-			if (!collected.cacheable) cacheable = false;
-			for (const entry of collected.entries) merged.set(entry.candidate.name, entry);
-		}
-		return {
-			entries: merged,
-			cacheable
-		};
-	}
-	async collectLayer(layer, options) {
-		const collected = await this.listLayerCandidates(layer, options);
-		collected.entries.sort(compareIndexedCandidates);
-		const seen = /* @__PURE__ */ new Set();
-		const result = [];
-		for (const entry of collected.entries) {
-			const skill = entry.candidate;
-			if (seen.has(skill.name)) {
-				this.ctx.logger.warn(`skill "${skill.name}" from ${skill.source} ignored because a higher-priority skill already exists`);
-				continue;
-			}
-			seen.add(skill.name);
-			result.push(entry);
-		}
-		return {
-			entries: result,
-			cacheable: collected.cacheable
-		};
-	}
-	async listLayerCandidates(layer, options) {
-		throwIfAborted(options.signal);
-		const candidates = [];
-		let cacheable = true;
-		let runtimeOrder = 0;
-		for (const skill of [...layer.runtime.values()].sort((a, b) => compareCodePoints(a.name, b.name))) {
-			candidates.push({
-				candidate: runtimeCandidate(skill),
-				provider: RUNTIME_SKILL_PROVIDER,
-				providerOrder: -1,
-				localOrder: runtimeOrder,
-				layer
-			});
-			runtimeOrder += 1;
-		}
-		for (const { provider, order } of [...layer.providers.values()]) {
-			let localOrder = 0;
-			let output;
-			try {
-				output = await waitWithAbort(provider.list(options), options.signal);
-			} catch (error) {
-				if (options.signal?.aborted === true) throw toError(options.signal.reason);
-				cacheable = false;
-				this.ctx.logger.warn(`skill provider "${provider.name}" skipped: ${errorMessage(error)}`);
-			}
-			if (output === void 0) continue;
-			const observation = normalizeProviderObservation(output, provider.name);
-			if (!observation.complete) cacheable = false;
-			for (const candidate of observation.candidates) {
-				validateCandidate(candidate, provider.name);
-				candidates.push({
-					candidate,
-					provider,
-					providerOrder: order,
-					localOrder,
-					layer
-				});
-				localOrder += 1;
-			}
-		}
-		return {
-			entries: candidates,
-			cacheable
-		};
-	}
-	invalidateCache() {
-		this.revision += 1;
-		this.collectCache.clear();
-		this.notifyChange();
-	}
-	/** Invalidate after a stale definition load, only while the exact registration that produced the entry is still live. */
-	invalidateEntry(entry) {
-		/* v8 ignore else -- A definition load can outlive the exact provider registration it selected. */
-		if (entry.layer.providers.get(entry.provider.name)?.provider === entry.provider) this.invalidateCache();
-	}
-	scopeId(key) {
-		let id = this.scopeIds.get(key);
-		if (id === void 0) {
-			id = this.nextScopeId;
-			this.nextScopeId += 1;
-			this.scopeIds.set(key, id);
-		}
-		return id;
-	}
-	collectCacheKey(cwd, chain, revision) {
-		return JSON.stringify({
-			cwd,
-			scopes: chain.map((key) => this.scopeId(key)),
-			revision
-		});
-	}
-	/** Notify catalog observers without making their refresh work load-bearing. */
-	notifyChange() {
-		for (const callback of this.ctx.events.dispatch("emit", ["skills/change"])) try {
-			const returned = callback();
-			Promise.resolve(returned).catch((error) => {
-				this.ctx.logger.warn(`skills/change listener rejected: ${errorMessage(error)}`);
-			});
-		} catch (error) {
-			this.ctx.logger.warn(`skills/change listener threw: ${errorMessage(error)}`);
-		}
-	}
-});
-function normalizeProviderObservation(output, providerName) {
-	if (Array.isArray(output)) return {
-		candidates: output,
-		complete: true
-	};
-	if (output === null || typeof output !== "object") throw invalidProviderObservation(providerName);
-	const observation = output;
-	if (!Array.isArray(observation.candidates) || typeof observation.complete !== "boolean") throw invalidProviderObservation(providerName);
-	return observation;
-}
-function invalidProviderObservation(providerName) {
-	return /* @__PURE__ */ new TypeError(`skill provider "${providerName}" list() must return an array or { candidates, complete } observation`);
-}
-const RUNTIME_SKILL_PROVIDER = {
-	name: RUNTIME_PROVIDER,
-	/* v8 ignore next -- Runtime skills are injected directly by the registry; this provider only owns `get()`. */
-	list() {
-		return Promise.resolve([]);
-	},
-	get(candidate) {
-		return Promise.resolve(candidate.locator);
-	}
-};
-function runtimeCandidate(skill) {
-	return {
-		name: skill.name,
-		description: skill.description,
-		...skill.whenToUse !== void 0 ? { whenToUse: skill.whenToUse } : {},
-		invocation: skill.invocation,
-		source: skill.source,
-		provider: skill.provider,
-		...skill.resourceBase !== void 0 ? { resourceBase: skill.resourceBase } : {},
-		rank: RUNTIME_RANK,
-		locator: skill,
-		...skill.path !== void 0 ? { path: skill.path } : {},
-		...skill.metadata !== void 0 ? { metadata: skill.metadata } : {}
-	};
-}
-function validateCandidate(candidate, providerName) {
-	if (typeof candidate.name !== "string") throw new TypeError(`skill provider "${providerName}" returned a non-string skill name`);
-	if (!SKILL_NAME.test(candidate.name)) throw new Error(`skill provider "${providerName}" returned invalid skill name "${candidate.name}"`);
-	if (typeof candidate.description !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string description`);
-	if (candidate.description.length === 0) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" without a description`);
-	validateInvocation(candidate.invocation, `skill provider "${providerName}" returned skill "${candidate.name}"`);
-	if (candidate.whenToUse !== void 0 && typeof candidate.whenToUse !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string whenToUse`);
-	if (typeof candidate.source !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string source`);
-	if (typeof candidate.rank !== "number" || !Number.isFinite(candidate.rank)) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" with an invalid rank`);
-	if (typeof candidate.provider !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string provider`);
-	if (candidate.provider !== providerName) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" for provider "${candidate.provider}"`);
-	if (candidate.path !== void 0 && typeof candidate.path !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string path`);
-}
-function validateRuntimeSkill(skill) {
-	if (!SKILL_NAME.test(skill.name)) throw new Error(`invalid skill name "${skill.name}"`);
-	if (skill.description.length === 0) throw new Error(`skill "${skill.name}" requires a description`);
-	validateInvocation(skill.invocation, `runtime skill "${skill.name}"`);
-}
-/** Validate a definition loaded from a provider-controlled parser or remote source. */
-function validateDefinition(skill) {
-	const name = skill.name;
-	const description = skill.description;
-	const whenToUse = skill.whenToUse;
-	const invocation = skill.invocation;
-	const source = skill.source;
-	const provider = skill.provider;
-	const content = skill.content;
-	const path = skill.path;
-	if (typeof name !== "string") throw new TypeError("loaded skill name must be a string");
-	if (!SKILL_NAME.test(name)) throw new Error(`loaded skill has invalid name "${name}"`);
-	if (typeof description !== "string") throw new TypeError(`loaded skill "${name}" description must be a string`);
-	if (description.length === 0) throw new Error(`loaded skill "${name}" requires a description`);
-	validateInvocation(invocation, `loaded skill "${name}"`);
-	if (whenToUse !== void 0 && typeof whenToUse !== "string") throw new TypeError(`loaded skill "${name}" whenToUse must be a string`);
-	if (typeof source !== "string") throw new TypeError(`loaded skill "${name}" source must be a string`);
-	if (typeof provider !== "string") throw new TypeError(`loaded skill "${name}" provider must be a string`);
-	if (typeof content !== "string") throw new TypeError(`loaded skill "${name}" content must be a string`);
-	if (path !== void 0 && typeof path !== "string") throw new TypeError(`loaded skill "${name}" path must be a string`);
-}
-function toSummary(skill) {
-	const { name, description, whenToUse, invocation, source, provider, resourceBase } = skill;
-	return {
-		name,
-		description,
-		...whenToUse !== void 0 ? { whenToUse } : {},
-		invocation,
-		source,
-		provider,
-		...resourceBase !== void 0 ? { resourceBase } : {}
-	};
-}
-function validateInvocation(invocation, subject) {
-	if (invocation === void 0) return;
-	if (typeof invocation !== "object" || invocation === null || Array.isArray(invocation)) throw new TypeError(`${subject} with a non-object invocation policy`);
-	const policy = invocation;
-	if (typeof policy.modelInvocable !== "boolean") throw new TypeError(`${subject} with a non-boolean invocation.modelInvocable`);
-	if (typeof policy.userInvocable !== "boolean") throw new TypeError(`${subject} with a non-boolean invocation.userInvocable`);
-}
-function compareSkillSummary(left, right) {
-	return compareCodePoints(left.name, right.name);
-}
-function compareCodePoints(left, right) {
-	if (left < right) return -1;
-	if (left > right) return 1;
-	return 0;
-}
-function compareIndexedCandidates(left, right) {
-	return left.candidate.rank - right.candidate.rank || left.providerOrder - right.providerOrder || left.localOrder - right.localOrder;
-}
-function assertPositiveInteger(name, value, minimum = 1) {
-	if (!Number.isInteger(value) || value < minimum) throw new Error(`skill: ${name} must be an integer greater than or equal to ${minimum}`);
-}
-function waitWithAbort(promise, signal) {
-	if (signal === void 0) return promise;
-	throwIfAborted(signal);
-	return new Promise((resolve, reject) => {
-		const cleanup = () => {
-			signal.removeEventListener("abort", onAbort);
-		};
-		const onAbort = () => {
-			cleanup();
-			reject(toError(signal.reason));
-		};
-		signal.addEventListener("abort", onAbort, { once: true });
-		promise.then((value) => {
-			cleanup();
-			resolve(value);
-		}, (error) => {
-			cleanup();
-			reject(toError(error));
-		});
-	});
-}
-/** Throw a total Error for an already-aborted lookup. */
-function throwIfAborted(signal) {
-	if (signal?.aborted === true) throw toError(signal.reason);
-}
-/** Normalize an arbitrary abort or provider failure without trusting coercion. */
-function toError(error) {
-	try {
-		if (error instanceof Error) return error;
-	} catch {}
-	return new Error(errorMessage(error));
-}
-/** Render an arbitrary provider failure without letting coercion escape containment. */
-function errorMessage(error) {
-	try {
-		return String(error);
-	} catch {
-		return "[unrenderable thrown value]";
-	}
-}
-//#endregion
-//#region lib/types/content.js
-/**
-* Ponytail skill bodies, ported from github.com/DietrichGebert/ponytail and
-* lightly adapted to the DeepSeek Harness surface (slash commands and the
-* `skill` tool). The `ponytail` skill is a mode-aware pointer card: the actual
-* ruleset is injected per session as the mode-filtered `PONYTAIL MODE ACTIVE`
-* section (see `instructions.ts`) and must not be duplicated here. The other
-* five skills ship verbatim as runtime skills.
-*
-* @module @deepseek-ai/dsh-ponytail
-*/
-/** The always-on lazy-senior-dev ruleset: also registered as a loadable skill. */
-const PONYTAIL_SKILL_BODY = `
-You are the ponytail persona — the lazy senior developer. Your active ruleset
-is ALREADY injected every turn as the "PONYTAIL MODE ACTIVE — level: <mode>"
-system-prompt section, filtered to this session's intensity. Follow exactly
-that section; do NOT reload, replace, or re-derive the ruleset from anywhere
-else — the section is the single source of truth and it is mode-aware.
-
-- Switch level: \`/ponytail lite|full|ultra|off\` (session-scoped)
-- Query: \`/ponytail status\`
-- Deactivate: "stop ponytail" / "normal mode"
-- One-shot skills: \`/ponytail-review\`, \`/ponytail-audit\`, \`/ponytail-debt\`,
-  \`/ponytail-gain\`, \`/ponytail-help\`
-- Reference: https://github.com/DietrichGebert/ponytail
-`;
-const PONYTAIL_DESCRIPTION = "Ponytail activation, modes, configuration, and help reference. The active ruleset is injected every turn by the system prompt; this skill is a pointer card. Use only when the user asks about Ponytail activation, modes, configuration, or help. Coding tasks already receive the active ruleset from the system prompt.";
-const REVIEW_SKILL_BODY = `
-Review diffs for unnecessary complexity. One line per finding: location, what
-to cut, what replaces it. The diff's best outcome is getting shorter.
+import { createRequire } from "node:module";
+import { Service } from "@deepseek-ai/cordis";
+import z from "@deepseek-ai/schemastery";
+import { mkdirSync, readFileSync, renameSync, unlinkSync, unwatchFile, watchFile, writeFileSync } from "node:fs";
+import { homedir } from "node:os";
+import { dirname, join } from "node:path";
+//#region ../../llm/llm/src/brand.ts
+/**
+* Brand a message identifier.
+* @param id - the opaque message identifier.
+* @returns the same string, branded; no validation is performed.
+*/
+function MessageId(id) {
+	return id;
+}
+//#endregion
+//#region ../../llm/llm/src/call-config.ts
+/**
+* Deep-freeze a value in place with an iterative traversal, guarding cycles,
+* so later mutation throws without imposing a JavaScript call-stack depth cap.
+* {@link AbortSignal} objects are deliberately skipped because they are the
+* request's live cancellation channel and freezing them breaks abort.
+* @param value - the value to freeze in place.
+* @returns the same value, frozen.
+*/
+function deepFreeze(value) {
+	const seen = /* @__PURE__ */ new WeakSet();
+	const pending = [{
+		kind: "visit",
+		node: value
+	}];
+	while (pending.length > 0) {
+		const task = pending.pop();
+		/* v8 ignore next -- the loop condition guarantees one pending task. */
+		if (task === void 0) continue;
+		if (task.kind === "property") {
+			pending.push({
+				kind: "visit",
+				node: task.source[task.key]
+			});
+			continue;
+		}
+		const node = task.node;
+		if (node === null || typeof node !== "object") continue;
+		if (node instanceof AbortSignal) continue;
+		if (seen.has(node)) continue;
+		seen.add(node);
+		Object.freeze(node);
+		const keys = Object.keys(node);
+		for (let index = keys.length - 1; index >= 0; index--) {
+			const key = keys[index];
+			/* v8 ignore next -- the loop is bounded by the captured key count. */
+			if (key === void 0) continue;
+			pending.push({
+				kind: "property",
+				source: node,
+				key
+			});
+		}
+	}
+	return value;
+}
+//#endregion
+//#region ../../llm/llm/src/message.ts
+/** Message value types, identity, and immutable construction helpers. */
+/**
+* Detach and deep-freeze a message whose identity already exists.
+* @param message - complete message, including its stable identity.
+* @returns an immutable snapshot that preserves the identity.
+*/
+function freezeMessage(message) {
+	return deepFreeze(structuredClone(message));
+}
+/**
+* Create one identified message and freeze it before publication.
+* @param input - complete role, content, and source for a new message.
+* @returns an immutable message with a fresh stable identity.
+*/
+function createMessage(input) {
+	return freezeMessage({
+		...input,
+		id: MessageId(crypto.randomUUID())
+	});
+}
+/**
+* Create one identified user-role message and freeze it before publication.
+* @param input - complete content and source for a new user message.
+* @returns an immutable user message with a fresh stable identity.
+*/
+function createUserMessage(input) {
+	return createMessage({
+		...input,
+		role: "user"
+	});
+}
+//#endregion
+//#region ../../util/timeout/src/index.ts
+/** Largest delay Node schedules without clamping it to one millisecond. */
+const MAX_TIMER_DELAY_MS = 2147483647;
+//#endregion
+//#region ../../llm/llm/src/error.ts
+/**
+* Canonical provider-neutral code for a response that completed normally but
+* carried no content blocks at all. Providers occasionally emit a degenerate
+* completion (a terminal stop with zero output); adapters classify it as this
+* failure instead of yielding an empty assistant message, because an empty
+* message silently ends the turn with nothing for the user or the loop to act
+* on. The attempt produced nothing durable, so retry policy treats it as safe
+* to repeat.
+*/
+const EMPTY_RESPONSE_CODE = "EMPTY_RESPONSE";
+new RegExp(String.raw`(?:^|[^a-z0-9])context[\s_-](?:length|window)[\s_-]` + String.raw`(?:exceed(?:ed|s)?|overflow(?:ed)?|limit[\s_-]exceeded)(?:$|[^a-z0-9])`, "i");
+new RegExp(String.raw`\b(?:request|prompt|input|messages?)\s+(?:is\s+|are\s+)?` + String.raw`too\s+(?:large|long)\s+for\s+(?:(?:this|the)\s+)?` + String.raw`(?:model(?:'s)?\s+)?context(?:\s+window)?\b`, "i");
+new RegExp(String.raw`\b(?:input|prompt|request|messages?)\b.{0,40}` + String.raw`\b(?:exceed(?:s|ed)?|overflows?|is\s+larger\s+than)\b.{0,40}` + String.raw`\b(?:the\s+)?(?:model(?:'s)?\s+)?context(?:\s+(?:length|window))?\b`, "i");
+//#endregion
+//#region ../../llm/llm/src/retry-policy.ts
+/**
+* Provider-owned request-retry policy configuration and resolution.
+*
+* Adapters expose one resolved policy per registered provider route; the
+* optional dsh-llm-retry plugin executes it on the agent's failed-step extension point.
+*
+* @module @deepseek-ai/dsh-llm/retry-policy
+*/
+const DEFAULT_MAX_RETRIES = 5;
+const DEFAULT_INITIAL_DELAY_MS = 500;
+const DEFAULT_MAX_DELAY_MS = 1e4;
+const DEFAULT_JITTER_RATIO = .1;
+const DEFAULT_RETRYABLE_CODES = Object.freeze([
+	EMPTY_RESPONSE_CODE,
+	"RATE_LIMIT",
+	"SERVER",
+	"TIMEOUT",
+	"TRANSPORT"
+]);
+const backoffSchema = z.object({
+	initialDelayMs: z.number().max(MAX_TIMER_DELAY_MS).default(DEFAULT_INITIAL_DELAY_MS),
+	maxDelayMs: z.number().max(MAX_TIMER_DELAY_MS).default(DEFAULT_MAX_DELAY_MS),
+	jitterRatio: z.number().min(0).max(1).default(DEFAULT_JITTER_RATIO)
+});
+const normalPolicySchema = z.object({
+	mode: z.const("normal").required(),
+	maxRetries: z.number().step(1).min(0).max(Number.MAX_SAFE_INTEGER).default(DEFAULT_MAX_RETRIES),
+	retryableCodes: z.array(z.string()).default([...DEFAULT_RETRYABLE_CODES]),
+	backoff: backoffSchema
+});
+const alwaysPolicySchema = z.object({
+	mode: z.const("always").required(),
+	backoff: backoffSchema
+});
+z.union([normalPolicySchema, alwaysPolicySchema]);
+//#endregion
+//#region ../../llm/llm/src/attribution.ts
+/**
+* Centralize the non-secret product identity every provider request sends as `User-Agent`, keeping
+* adapters from drifting. See
+* `.agents/notes/implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md`.
+*
+* App-attribution vocabulary for provider requests.
+* @module @deepseek-ai/dsh-llm/attribution
+*/
+const { version } = createRequire(import.meta.url)("../package.json");
+//#endregion
+//#region ../../llm/llm/src/never.ts
+/**
+* Exhaustiveness helper for closed core unions. Use {@link assertNever} at the default branch so a
+* new variant fails compilation at every required handler. Do not use it for declaration-merged
+* unions such as session events or content blocks: handle known variants and explicitly fall
+* through because plugins may add valid unknown cases.
+* @module @deepseek-ai/dsh-llm/never
+*/
+/**
+* Mark an unreachable closed-union branch. A newly unhandled typed variant fails at the call site;
+* a value that escaped its type throws with diagnostics at runtime.
+* @param value - the impossible value; typed `never` so an unhandled variant fails compilation at the call site.
+* @param context - optional label (e.g. the switch site) prefixed into the throw message.
+* @returns never — it always throws, with the offending value JSON-rendered in the message.
+*/
+function assertNever(value, context) {
+	const rendered = JSON.stringify(value) ?? String(value);
+	throw new Error(`unreachable variant${context ? ` in ${context}` : ""}: ${rendered}`);
+}
+//#endregion
+//#region ../../core/scope/src/store.ts
+/**
+* Insertion-ordered named entries with caller-owned duplicate diagnostics.
+*
+* Values are borrowed. Iterators are live within one nonempty table
+* generation; draining the table detaches them from later insertions. Each
+* successful insertion returns an idempotent undo for that exact entry.
+*/
+var NamedEntries = class {
+	duplicateError;
+	data = /* @__PURE__ */ new Map();
+	constructor(duplicateError) {
+		this.duplicateError = duplicateError;
+	}
+	/**
+	* Insert one unique name.
+	* @param name - name unique within this table.
+	* @param value - borrowed value to retain.
+	* @returns an idempotent undo that removes only this insertion.
+	*/
+	insert(name, value) {
+		const data = this.data;
+		if (data.has(name)) throw this.duplicateError(name);
+		data.set(name, value);
+		let active = true;
+		return () => {
+			if (!active) return;
+			active = false;
+			data.delete(name);
+			if (data.size === 0 && this.data === data) this.data = /* @__PURE__ */ new Map();
+		};
+	}
+	/**
+	* Read one named value.
+	* @param name - name to resolve.
+	* @returns the retained value, or `undefined` when absent.
+	*/
+	get(name) {
+		return this.data.get(name);
+	}
+	/**
+	* Test one name for membership.
+	* @param name - name to test.
+	* @returns whether the table contains that name.
+	*/
+	has(name) {
+		return this.data.has(name);
+	}
+	/**
+	* Iterate live names in insertion order.
+	* @returns the native live key iterator.
+	*/
+	keys() {
+		return this.data.keys();
+	}
+	/**
+	* Iterate live entries in insertion order.
+	* @returns the native live entry iterator.
+	*/
+	entries() {
+		return this.data.entries();
+	}
+	/**
+	* Iterate live values in insertion order.
+	* @returns the native live value iterator.
+	*/
+	values() {
+		return this.data.values();
+	}
+	/**
+	* Test whether this table has no entries.
+	* @returns whether the table is empty.
+	*/
+	isEmpty() {
+		return this.data.size === 0;
+	}
+};
+/**
+* Own the global and exact-scope layers for one registry.
+*
+* Reads never create scoped layers. Registrations derive both visibility and
+* effect ownership from the supplied Cordis context, collect undo before
+* notification, and reclaim only a completely empty aggregate layer.
+*/
+var ScopedLayers = class {
+	createLayer;
+	onChange;
+	/** The eagerly constructed context-global layer. */
+	global;
+	scoped = /* @__PURE__ */ new Map();
+	constructor(createLayer, onChange) {
+		this.createLayer = createLayer;
+		this.onChange = onChange;
+		this.global = createLayer(void 0);
+	}
+	/**
+	* Read an existing exact-scope overlay. Deliberately chain-blind: callers
+	* addressing one scope's OWN contributions (its restrictions, its guards)
+	* must not silently pick up an ancestor's — use {@link chainLayers} where
+	* inheritance is the point.
+	* @param scope - exact scope key; `undefined` denotes no overlay.
+	* @returns the existing scoped layer, or `undefined` without creating one.
+	*/
+	peek(scope) {
+		if (scope === void 0) return void 0;
+		return this.scoped.get(scope);
+	}
+	/**
+	* Existing overlays along the scope's parent chain ({@link scopeChainOf}),
+	* farthest ancestor first and the exact scope last, so a caller layering
+	* them in order gives the nearest scope the final word.
+	* @param scope - viewing scope, or `undefined` for no overlays.
+	* @returns the existing layers, nearest last; absent overlays are skipped.
+	*/
+	chainLayers(scope) {
+		const layers = [];
+		for (const key of scopeChainOf(scope).reverse()) {
+			const layer = this.scoped.get(key);
+			if (layer !== void 0) layers.push(layer);
+		}
+		return layers;
+	}
+	/**
+	* Materialize global named entries followed by scope-chain shadows,
+	* farthest ancestor first, so the nearest scope's entry wins a name.
+	* @param scope - viewing scope, or `undefined` for the global view.
+	* @param pick - select the named table from a layer.
+	* @returns an insertion-ordered effective map.
+	*/
+	merge(scope, pick) {
+		const merged = new Map(pick(this.global).entries());
+		for (const layer of this.chainLayers(scope)) for (const [name, value] of pick(layer).entries()) merged.set(name, value);
+		return merged;
+	}
+	/**
+	* Attach one synchronous layer mutation to its registration context.
+	* @param ctx - context that determines both scope visibility and effect ownership.
+	* @param action - atomic mutation returning its synchronous undo.
+	* @param options - Cordis effect label and optional change notification.
+	* @returns the exact disposer returned by `ctx.effect()`.
+	*/
+	effect(ctx, action, options) {
+		const scope = scopeOf(ctx);
+		const notify = options.notify ?? true;
+		return ctx.effect(function* () {
+			let layer;
+			let created = false;
+			if (scope === void 0) layer = this.global;
+			else {
+				const existing = this.scoped.get(scope);
+				if (existing === void 0) {
+					layer = this.createLayer(scope);
+					this.scoped.set(scope, layer);
+					created = true;
+				} else layer = existing;
+			}
+			let undo;
+			try {
+				undo = action(layer);
+			} catch (error) {
+				if (scope !== void 0 && created && layer.isEmpty()) this.scoped.delete(scope);
+				throw error;
+			}
+			yield () => {
+				undo();
+				if (scope !== void 0 && layer.isEmpty()) this.scoped.delete(scope);
+				if (notify) this.onChange();
+			};
+			if (notify) this.onChange();
+		}.bind(this), options.label);
+	}
+};
+//#endregion
+//#region ../../core/scope/src/index.ts
+/** Context tag written by {@link createScope}. */
+const kScope = Symbol("dsh.scope");
+/**
+* The enclosing scope of each key. One relation powers both directions of
+* scope nesting: registration views inherit DOWN the chain (a child scope
+* sees its ancestors' layers — {@link ScopedLayers}), and event admission
+* extends UP it (a listener tagged with an ancestor receives events dispatched
+* to a descendant key — {@link scopeTarget}).
+*/
+const scopeParents = /* @__PURE__ */ new WeakMap();
+/**
+* The chain from a key to its root ancestor.
+* @param key - the starting key, or `undefined` for the empty chain.
+* @returns keys nearest-first: `[key, parent, grandparent, …]`.
+*/
+function scopeChainOf(key) {
+	const chain = [];
+	for (let cursor = key; cursor !== void 0; cursor = scopeParents.get(cursor)) chain.push(cursor);
+	return chain;
+}
+/**
+* Read the nearest scope tag inherited by a context.
+* @param ctx - context to inspect.
+* @returns its scope key, or `undefined` for an unscoped context.
+*/
+function scopeOf(ctx) {
+	return ctx[kScope];
+}
+//#endregion
+//#region ../../skill/skill/src/index.ts
+/**
+* Agent skill provider registry.
+*
+* This package owns the Service Definition role of the skill capability seam.
+* Concrete
+* providers such as `@deepseek-ai/dsh-skill-filesystem` decide where skills come
+* from; this service only merges provider catalogs, resolves the winning skill
+* for a name, and exposes the winning summaries and definitions to consumers.
+*
+* @module @deepseek-ai/dsh-skill
+*/
+const SKILL_NAME = /^[a-z0-9]+(?:-[a-z0-9]+)*$/;
+const DEFAULT_COLLECT_CACHE_ENTRIES = 128;
+const MAX_COLLECT_ATTEMPTS = 2;
+const RUNTIME_PROVIDER = "runtime";
+const RUNTIME_RANK = 250;
+/**
+* Return whether a string is a valid kebab-case skill name.
+* @param name - candidate skill name to validate.
+* @returns whether the name matches the public skill-name grammar.
+*/
+function isSkillName(name) {
+	return SKILL_NAME.test(name);
+}
+/**
+* Render one loaded skill for the model. The output is shared verbatim by the
+* `skill` tool result and the user-explicit invocation injection, so the model
+* sees one canonical `<skill_content>` shape on both paths. The name rides an
+* escaped attribute; the body is embedded verbatim (skills are trusted local
+* content, and user-supplied invocation text stays outside this wrapper).
+* @param skill - name, provider, optional resource base, and body to render.
+* @returns the complete model-facing `<skill_content>` block.
+*/
+function renderSkillContent(skill) {
+	const resourceHint = renderResourceHint(skill);
+	return [
+		`<skill_content name="${escapeAttr(skill.name)}">`,
+		"<skill_resources>",
+		...resourceHint,
+		"</skill_resources>",
+		"",
+		"<skill_instructions>",
+		skill.content,
+		"</skill_instructions>",
+		"</skill_content>"
+	].join("\n");
+}
+function renderResourceHint(skill) {
+	const base = skill.resourceBase;
+	if (base === void 0) return [`Resources for this skill are managed by provider "${escapeText(skill.provider)}".`, "Load referenced resources only as needed."];
+	switch (base.kind) {
+		case "directory": return [`Base directory for this skill: ${escapeText(base.path)}`, "Resolve relative paths mentioned by this skill against the base directory before using them. Load referenced resources only as needed."];
+		case "url": return [`Base URL for this skill: ${escapeText(base.url)}`, "Resolve relative URLs mentioned by this skill against the base URL before using them. Load referenced resources only as needed."];
+		case "opaque": return [`Resources for this skill: ${escapeText(base.description)}`, "Load referenced resources only as needed."];
+		/* v8 ignore start -- SkillResourceBase is a closed union; a future kind must fail compilation here. */
+		default: return assertNever(base, "SkillResourceBase.kind");
+	}
+}
+function escapeAttr(value) {
+	return value.replaceAll("&", "&amp;").replaceAll("\"", "&quot;").replaceAll("<", "&lt;");
+}
+/**
+* Escape model-facing prose embedded inside skill markup so provider-supplied
+* text cannot open or close framing tags.
+* @param value - raw prose to embed.
+* @returns the escaped text.
+*/
+function escapeText(value) {
+	return value.replaceAll("&", "&amp;").replaceAll("<", "&lt;").replaceAll(">", "&gt;");
+}
+/** One scope's complete skill-registry contribution. */
+var SkillLayer = class {
+	/** Providers registered through contexts carrying this scope, insertion-ordered. */
+	providers;
+	/** Runtime skills registered through contexts carrying this scope. */
+	runtime = /* @__PURE__ */ new Map();
+	constructor(scope) {
+		this.providers = new NamedEntries((name) => /* @__PURE__ */ new Error(scope === void 0 ? `a skill provider named "${name}" is already registered` : `a skill provider named "${name}" is already registered in this scope`));
+	}
+	/** Whether every contribution table in this aggregate layer is empty. */
+	isEmpty() {
+		return this.providers.isEmpty() && this.runtime.size === 0;
+	}
+};
+(class extends Service {
+	static Config = z.object({ collectCacheMaxEntries: z.number().default(DEFAULT_COLLECT_CACHE_ENTRIES) });
+	collectCacheMaxEntries;
+	layers = new ScopedLayers((scope) => new SkillLayer(scope), () => {
+		this.invalidateCache();
+	});
+	collectCache = /* @__PURE__ */ new Map();
+	revision = 0;
+	nextProviderOrder = 0;
+	/** Stable identities for cache keys; scope keys are opaque identity-compared objects. */
+	scopeIds = /* @__PURE__ */ new WeakMap();
+	nextScopeId = 1;
+	constructor(ctx, config = {}) {
+		super(ctx, "skills");
+		this.collectCacheMaxEntries = config.collectCacheMaxEntries ?? DEFAULT_COLLECT_CACHE_ENTRIES;
+		assertPositiveInteger("collectCacheMaxEntries", this.collectCacheMaxEntries);
+	}
+	/**
+	* Register a borrowed same-process provider synchronously during plugin
+	* apply, into the calling context's layer: a scoped context (an agent
+	* preset's standing mount) registers for that scope alone, an unscoped
+	* context registers globally. Duplicate names within one layer and reserved
+	* names throw; remote initialization belongs in `list()`. Fiber disposal
+	* unregisters the provider and invalidates catalog caches.
+	* @param create - synchronous factory receiving this registration's lifecycle and invalidation control.
+	* @returns the exact Cordis effect disposer that unregisters this provider;
+	*   composite effects may yield it directly to preserve teardown ordering.
+	*/
+	registerProvider(create) {
+		const lifecycle = new AbortController();
+		let registration;
+		let provider;
+		const control = {
+			signal: lifecycle.signal,
+			invalidate: () => {
+				const active = registration;
+				if (active !== void 0 && active.layer.providers.get(active.name)?.provider === provider) this.invalidateCache();
+			}
+		};
+		try {
+			provider = create(control);
+			const name = provider.name;
+			if (name === RUNTIME_PROVIDER) throw new Error(`"${RUNTIME_PROVIDER}" is reserved for runtime skill registrations`);
+			const order = this.nextProviderOrder;
+			this.nextProviderOrder += 1;
+			return this.layers.effect(this.ctx, (layer) => {
+				const undo = layer.providers.insert(name, {
+					provider,
+					order
+				});
+				registration = {
+					layer,
+					name
+				};
+				return () => {
+					registration = void 0;
+					undo();
+					lifecycle.abort(/* @__PURE__ */ new Error(`skill provider "${name}" disposed`));
+				};
+			}, { label: "skills.registerProvider()" });
+		} catch (error) {
+			lifecycle.abort(error);
+			throw error;
+		}
+	}
+	/**
+	* Register a borrowed readonly runtime skill into the calling context's
+	* layer. Project entries outrank runtime entries, which outrank user
+	* entries, within one layer. Same-name runtime entries in one layer are
+	* first-wins; a duplicate logs a warning and receives a no-op disposer so
+	* it cannot remove the winner.
+	* @param skill - the skill definition input; omitted invocation and provider fields receive defaults.
+	* @returns the exact Cordis effect disposer, preserving composite teardown order and invalidating caches.
+	*/
+	register(skill) {
+		validateRuntimeSkill(skill);
+		const scope = scopeOf(this.ctx);
+		const existingLayer = scope === void 0 ? this.layers.global : this.layers.peek(scope);
+		if (existingLayer !== void 0 && existingLayer.runtime.has(skill.name)) {
+			this.ctx.logger.warn(`runtime skill "${skill.name}" ignored because it is already registered`);
+			return () => {};
+		}
+		const definition = {
+			...skill,
+			invocation: skill.invocation ?? {
+				modelInvocable: true,
+				userInvocable: true
+			},
+			provider: skill.provider ?? RUNTIME_PROVIDER
+		};
+		return this.layers.effect(this.ctx, (layer) => {
+			layer.runtime.set(definition.name, definition);
+			return () => {
+				layer.runtime.delete(definition.name);
+			};
+		}, { label: "skills.register()" });
+	}
+	/**
+	* List invocation-neutral skill summaries for a workspace. Consumers apply
+	* model or user invocation policy at their operational boundary. Lookup
+	* options and provider candidates are readonly same-process values borrowed
+	* throughout discovery.
+	* @param options - view options; `scope` selects the viewing agent's layers, `cwd` selects project roots, and `signal` cancels discovery.
+	* @returns all sorted winning summaries.
+	*/
+	async list(options = {}) {
+		return (await this.snapshot(options)).skills;
+	}
+	/**
+	* Observe the current invocation-neutral catalog and whether discovery completed within a stable revision.
+	* Incomplete observations are never cached, allowing consumers to retain last-good state and
+	* retry on their next request boundary.
+	* @param options - view options; `scope` selects the viewing agent's layers, `cwd` selects project roots, and `signal` cancels discovery.
+	* @returns sorted summaries plus discovery-completeness state.
+	*/
+	async snapshot(options = {}) {
+		const collected = await this.collect(options);
+		return {
+			skills: [...collected.entries.values()].map((entry) => toSummary(entry.candidate)).sort(compareSkillSummary),
+			complete: collected.cacheable
+		};
+	}
+	/**
+	* Load and validate the winning candidate, passing its opaque discovery locator back to the
+	* provider. Cancellation is rechecked after selection, including cache hits, and raced against
+	* loading so an uncooperative provider cannot hang the caller.
+	* @param name - kebab-case skill name.
+	* @param options - view options; `scope` selects the viewing agent's layers,
+	*   `cwd` selects workspace-sensitive skills, and `signal` cancels work.
+	* @returns the full skill, including body content, or `undefined`.
+	*/
+	async get(name, options = {}) {
+		if (!isSkillName(name)) return void 0;
+		const collected = await this.collect(options);
+		throwIfAborted(options.signal);
+		const match = collected.entries.get(name);
+		if (match === void 0) return void 0;
+		const definition = await waitWithAbort(match.provider.get(match.candidate, options), options.signal);
+		if (definition === void 0) return void 0;
+		validateDefinition(definition);
+		if (definition.name !== match.candidate.name) {
+			this.invalidateEntry(match);
+			return;
+		}
+		return definition;
+	}
+	async collect(options) {
+		throwIfAborted(options.signal);
+		let attempt = 1;
+		while (true) {
+			const revision = this.revision;
+			const key = this.collectCacheKey(options.cwd, scopeChainOf(options.scope), revision);
+			const cached = this.collectCache.get(key);
+			if (cached !== void 0) return {
+				entries: cached,
+				cacheable: true
+			};
+			const result = await this.collectFresh(options);
+			throwIfAborted(options.signal);
+			if (revision !== this.revision) {
+				if (attempt < MAX_COLLECT_ATTEMPTS) {
+					attempt += 1;
+					continue;
+				}
+				return {
+					entries: result.entries,
+					cacheable: false
+				};
+			}
+			if (result.cacheable) {
+				this.collectCache.set(key, result.entries);
+				if (this.collectCache.size > this.collectCacheMaxEntries) {
+					const oldest = this.collectCache.keys().next();
+					this.collectCache.delete(oldest.value);
+				}
+			}
+			return result;
+		}
+	}
+	async collectFresh(options) {
+		const layers = [this.layers.global, ...this.layers.chainLayers(options.scope)];
+		const merged = /* @__PURE__ */ new Map();
+		let cacheable = true;
+		for (const layer of layers) {
+			const collected = await this.collectLayer(layer, options);
+			if (!collected.cacheable) cacheable = false;
+			for (const entry of collected.entries) merged.set(entry.candidate.name, entry);
+		}
+		return {
+			entries: merged,
+			cacheable
+		};
+	}
+	async collectLayer(layer, options) {
+		const collected = await this.listLayerCandidates(layer, options);
+		collected.entries.sort(compareIndexedCandidates);
+		const seen = /* @__PURE__ */ new Set();
+		const result = [];
+		for (const entry of collected.entries) {
+			const skill = entry.candidate;
+			if (seen.has(skill.name)) {
+				this.ctx.logger.warn(`skill "${skill.name}" from ${skill.source} ignored because a higher-priority skill already exists`);
+				continue;
+			}
+			seen.add(skill.name);
+			result.push(entry);
+		}
+		return {
+			entries: result,
+			cacheable: collected.cacheable
+		};
+	}
+	async listLayerCandidates(layer, options) {
+		throwIfAborted(options.signal);
+		const candidates = [];
+		let cacheable = true;
+		let runtimeOrder = 0;
+		for (const skill of [...layer.runtime.values()].sort((a, b) => compareCodePoints(a.name, b.name))) {
+			candidates.push({
+				candidate: runtimeCandidate(skill),
+				provider: RUNTIME_SKILL_PROVIDER,
+				providerOrder: -1,
+				localOrder: runtimeOrder,
+				layer
+			});
+			runtimeOrder += 1;
+		}
+		for (const { provider, order } of [...layer.providers.values()]) {
+			let localOrder = 0;
+			let output;
+			try {
+				output = await waitWithAbort(provider.list(options), options.signal);
+			} catch (error) {
+				if (options.signal?.aborted === true) throw toError(options.signal.reason);
+				cacheable = false;
+				this.ctx.logger.warn(`skill provider "${provider.name}" skipped: ${errorMessage(error)}`);
+			}
+			if (output === void 0) continue;
+			const observation = normalizeProviderObservation(output, provider.name);
+			if (!observation.complete) cacheable = false;
+			for (const candidate of observation.candidates) {
+				validateCandidate(candidate, provider.name);
+				candidates.push({
+					candidate,
+					provider,
+					providerOrder: order,
+					localOrder,
+					layer
+				});
+				localOrder += 1;
+			}
+		}
+		return {
+			entries: candidates,
+			cacheable
+		};
+	}
+	invalidateCache() {
+		this.revision += 1;
+		this.collectCache.clear();
+		this.notifyChange();
+	}
+	/** Invalidate after a stale definition load, only while the exact registration that produced the entry is still live. */
+	invalidateEntry(entry) {
+		/* v8 ignore else -- A definition load can outlive the exact provider registration it selected. */
+		if (entry.layer.providers.get(entry.provider.name)?.provider === entry.provider) this.invalidateCache();
+	}
+	scopeId(key) {
+		let id = this.scopeIds.get(key);
+		if (id === void 0) {
+			id = this.nextScopeId;
+			this.nextScopeId += 1;
+			this.scopeIds.set(key, id);
+		}
+		return id;
+	}
+	collectCacheKey(cwd, chain, revision) {
+		return JSON.stringify({
+			cwd,
+			scopes: chain.map((key) => this.scopeId(key)),
+			revision
+		});
+	}
+	/** Notify catalog observers without making their refresh work load-bearing. */
+	notifyChange() {
+		for (const callback of this.ctx.events.dispatch("emit", ["skills/change"])) try {
+			const returned = callback();
+			Promise.resolve(returned).catch((error) => {
+				this.ctx.logger.warn(`skills/change listener rejected: ${errorMessage(error)}`);
+			});
+		} catch (error) {
+			this.ctx.logger.warn(`skills/change listener threw: ${errorMessage(error)}`);
+		}
+	}
+});
+function normalizeProviderObservation(output, providerName) {
+	if (Array.isArray(output)) return {
+		candidates: output,
+		complete: true
+	};
+	if (output === null || typeof output !== "object") throw invalidProviderObservation(providerName);
+	const observation = output;
+	if (!Array.isArray(observation.candidates) || typeof observation.complete !== "boolean") throw invalidProviderObservation(providerName);
+	return observation;
+}
+function invalidProviderObservation(providerName) {
+	return /* @__PURE__ */ new TypeError(`skill provider "${providerName}" list() must return an array or { candidates, complete } observation`);
+}
+const RUNTIME_SKILL_PROVIDER = {
+	name: RUNTIME_PROVIDER,
+	/* v8 ignore next -- Runtime skills are injected directly by the registry; this provider only owns `get()`. */
+	list() {
+		return Promise.resolve([]);
+	},
+	get(candidate) {
+		return Promise.resolve(candidate.locator);
+	}
+};
+function runtimeCandidate(skill) {
+	return {
+		name: skill.name,
+		description: skill.description,
+		...skill.whenToUse !== void 0 ? { whenToUse: skill.whenToUse } : {},
+		invocation: skill.invocation,
+		source: skill.source,
+		provider: skill.provider,
+		...skill.resourceBase !== void 0 ? { resourceBase: skill.resourceBase } : {},
+		rank: RUNTIME_RANK,
+		locator: skill,
+		...skill.path !== void 0 ? { path: skill.path } : {},
+		...skill.metadata !== void 0 ? { metadata: skill.metadata } : {}
+	};
+}
+function validateCandidate(candidate, providerName) {
+	if (typeof candidate.name !== "string") throw new TypeError(`skill provider "${providerName}" returned a non-string skill name`);
+	if (!SKILL_NAME.test(candidate.name)) throw new Error(`skill provider "${providerName}" returned invalid skill name "${candidate.name}"`);
+	if (typeof candidate.description !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string description`);
+	if (candidate.description.length === 0) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" without a description`);
+	validateInvocation(candidate.invocation, `skill provider "${providerName}" returned skill "${candidate.name}"`);
+	if (candidate.whenToUse !== void 0 && typeof candidate.whenToUse !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string whenToUse`);
+	if (typeof candidate.source !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string source`);
+	if (typeof candidate.rank !== "number" || !Number.isFinite(candidate.rank)) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" with an invalid rank`);
+	if (typeof candidate.provider !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string provider`);
+	if (candidate.provider !== providerName) throw new Error(`skill provider "${providerName}" returned skill "${candidate.name}" for provider "${candidate.provider}"`);
+	if (candidate.path !== void 0 && typeof candidate.path !== "string") throw new TypeError(`skill provider "${providerName}" returned skill "${candidate.name}" with a non-string path`);
+}
+function validateRuntimeSkill(skill) {
+	if (!SKILL_NAME.test(skill.name)) throw new Error(`invalid skill name "${skill.name}"`);
+	if (skill.description.length === 0) throw new Error(`skill "${skill.name}" requires a description`);
+	validateInvocation(skill.invocation, `runtime skill "${skill.name}"`);
+}
+/** Validate a definition loaded from a provider-controlled parser or remote source. */
+function validateDefinition(skill) {
+	const name = skill.name;
+	const description = skill.description;
+	const whenToUse = skill.whenToUse;
+	const invocation = skill.invocation;
+	const source = skill.source;
+	const provider = skill.provider;
+	const content = skill.content;
+	const path = skill.path;
+	if (typeof name !== "string") throw new TypeError("loaded skill name must be a string");
+	if (!SKILL_NAME.test(name)) throw new Error(`loaded skill has invalid name "${name}"`);
+	if (typeof description !== "string") throw new TypeError(`loaded skill "${name}" description must be a string`);
+	if (description.length === 0) throw new Error(`loaded skill "${name}" requires a description`);
+	validateInvocation(invocation, `loaded skill "${name}"`);
+	if (whenToUse !== void 0 && typeof whenToUse !== "string") throw new TypeError(`loaded skill "${name}" whenToUse must be a string`);
+	if (typeof source !== "string") throw new TypeError(`loaded skill "${name}" source must be a string`);
+	if (typeof provider !== "string") throw new TypeError(`loaded skill "${name}" provider must be a string`);
+	if (typeof content !== "string") throw new TypeError(`loaded skill "${name}" content must be a string`);
+	if (path !== void 0 && typeof path !== "string") throw new TypeError(`loaded skill "${name}" path must be a string`);
+}
+function toSummary(skill) {
+	const { name, description, whenToUse, invocation, source, provider, resourceBase } = skill;
+	return {
+		name,
+		description,
+		...whenToUse !== void 0 ? { whenToUse } : {},
+		invocation,
+		source,
+		provider,
+		...resourceBase !== void 0 ? { resourceBase } : {}
+	};
+}
+function validateInvocation(invocation, subject) {
+	if (invocation === void 0) return;
+	if (typeof invocation !== "object" || invocation === null || Array.isArray(invocation)) throw new TypeError(`${subject} with a non-object invocation policy`);
+	const policy = invocation;
+	if (typeof policy.modelInvocable !== "boolean") throw new TypeError(`${subject} with a non-boolean invocation.modelInvocable`);
+	if (typeof policy.userInvocable !== "boolean") throw new TypeError(`${subject} with a non-boolean invocation.userInvocable`);
+}
+function compareSkillSummary(left, right) {
+	return compareCodePoints(left.name, right.name);
+}
+function compareCodePoints(left, right) {
+	if (left < right) return -1;
+	if (left > right) return 1;
+	return 0;
+}
+function compareIndexedCandidates(left, right) {
+	return left.candidate.rank - right.candidate.rank || left.providerOrder - right.providerOrder || left.localOrder - right.localOrder;
+}
+function assertPositiveInteger(name, value, minimum = 1) {
+	if (!Number.isInteger(value) || value < minimum) throw new Error(`skill: ${name} must be an integer greater than or equal to ${minimum}`);
+}
+function waitWithAbort(promise, signal) {
+	if (signal === void 0) return promise;
+	throwIfAborted(signal);
+	return new Promise((resolve, reject) => {
+		const cleanup = () => {
+			signal.removeEventListener("abort", onAbort);
+		};
+		const onAbort = () => {
+			cleanup();
+			reject(toError(signal.reason));
+		};
+		signal.addEventListener("abort", onAbort, { once: true });
+		promise.then((value) => {
+			cleanup();
+			resolve(value);
+		}, (error) => {
+			cleanup();
+			reject(toError(error));
+		});
+	});
+}
+/** Throw a total Error for an already-aborted lookup. */
+function throwIfAborted(signal) {
+	if (signal?.aborted === true) throw toError(signal.reason);
+}
+/** Normalize an arbitrary abort or provider failure without trusting coercion. */
+function toError(error) {
+	try {
+		if (error instanceof Error) return error;
+	} catch {}
+	return new Error(errorMessage(error));
+}
+/** Render an arbitrary provider failure without letting coercion escape containment. */
+function errorMessage(error) {
+	try {
+		return String(error);
+	} catch {
+		return "[unrenderable thrown value]";
+	}
+}
+//#endregion
+//#region lib/types/content.js
+/**
+* Ponytail skill bodies, ported from github.com/DietrichGebert/ponytail and
+* lightly adapted to the DeepSeek Harness surface (slash commands and the
+* `skill` tool). The `ponytail` skill is a mode-aware pointer card: the actual
+* ruleset is injected per session as the mode-filtered `PONYTAIL MODE ACTIVE`
+* section (see `instructions.ts`) and must not be duplicated here. The other
+* five skills ship verbatim as runtime skills.
+*
+* @module @mengyuly/dsh-ponytail
+*/
+/** The always-on lazy-senior-dev ruleset: also registered as a loadable skill. */
+const PONYTAIL_SKILL_BODY = `
+You are the ponytail persona — the lazy senior developer. Your active ruleset
+is ALREADY injected every turn as the "PONYTAIL MODE ACTIVE — level: <mode>"
+system-prompt section, filtered to this session's intensity. Follow exactly
+that section; do NOT reload, replace, or re-derive the ruleset from anywhere
+else — the section is the single source of truth and it is mode-aware.
+
+- Switch level: \`/ponytail lite|full|ultra|off\` (session-scoped)
+- Query: \`/ponytail status\`
+- Deactivate: "stop ponytail" / "normal mode"
+- One-shot skills: \`/ponytail-review\`, \`/ponytail-audit\`, \`/ponytail-debt\`,
+  \`/ponytail-gain\`, \`/ponytail-help\`
+- Reference: https://github.com/DietrichGebert/ponytail
+`;
+const PONYTAIL_DESCRIPTION = "Ponytail activation, modes, configuration, and help reference. The active ruleset is injected every turn by the system prompt; this skill is a pointer card. Use only when the user asks about Ponytail activation, modes, configuration, or help. Coding tasks already receive the active ruleset from the system prompt.";
+const REVIEW_SKILL_BODY = `
+Review the current diff for unnecessary complexity. Inspect affected callers
+and tests first. Report only when code evidence shows the replacement preserves
+required behavior.
+
+## Format
+
+\`<file>:L<start>-<end>: <tag> <what>. Replacement: <simpler form>. Evidence: <observable reason>.\`
+
+Tags:
+
+- \`delete:\` dead code, unused flexibility, speculative feature. Replacement: nothing.
+- \`stdlib:\` hand-rolled thing the standard library ships. Name the function.
+- \`native:\` dependency or code doing what the platform already does. Name the feature.
+- \`yagni:\` abstraction with one implementation, config nobody sets, layer with one caller.
+- \`shrink:\` same logic, fewer lines. Show the shorter form.
+
+Evidence must name an actual caller count, unused export, duplicate branch, or
+equivalent stdlib/native behavior. A name that merely looks abstract is not
+evidence. If preserving behavior is uncertain, omit the finding.
+
+## Scoring
+
+End with \`net: -<N> lines countable.\` only when concrete ranges make the total
+countable; otherwise say \`net: uncounted.\` Nothing to cut: \`Lean already. Ship.\`
+
+## Boundaries
+
+Scope: over-engineering and complexity only. Correctness bugs, security holes,
+and performance are explicitly out of scope. Route them to a normal review
+pass, not this one. A single smoke test or \`assert\`-based
+self-check is the ponytail minimum, not bloat, never flag it for deletion.
+Does not apply the fixes, only lists them.
+"stop ponytail-review" or "normal mode": revert to verbose review style.
+`;
+const REVIEW_DESCRIPTION = "Code review focused exclusively on over-engineering. Finds what to delete: reinvented standard library, unneeded dependencies, speculative abstractions, dead flexibility. One line per finding: location, what to cut, what replaces it. Use when the user says \"review for over-engineering\", \"what can we delete\", \"is this over-engineered\", \"simplify review\", or invokes /ponytail-review. Complements correctness-focused review, this one only hunts complexity.";
+const AUDIT_SKILL_BODY = `
+Audit repository complexity. Skip generated, vendored, dependency, and build
+output. Inspect manifests, callers, and tests. Return the top 10 findings.
 
 ## Format
 
-\`L<line>: <tag> <what>. <replacement>.\`, or \`<file>:L<line>: ...\` for
-multi-file diffs.
-
-Tags:
-
-- \`delete:\` dead code, unused flexibility, speculative feature. Replacement: nothing.
-- \`stdlib:\` hand-rolled thing the standard library ships. Name the function.
-- \`native:\` dependency or code doing what the platform already does. Name the feature.
-- \`yagni:\` abstraction with one implementation, config nobody sets, layer with one caller.
-- \`shrink:\` same logic, fewer lines. Show the shorter form.
-
-## Examples
-
-❌ "This EmailValidator class might be more complex than necessary, have you
-considered whether all these validation rules are needed at this stage?"
-
-✅ \`L12-38: stdlib: 27-line validator class. "@" in email, 1 line, real validation is the confirmation mail.\`
-
-✅ \`L4: native: moment.js imported for one format call. Intl.DateTimeFormat, 0 deps.\`
-
-✅ \`repo.py:L88: yagni: AbstractRepository with one implementation. Inline it until a second one exists.\`
-
-✅ \`L52-71: delete: retry wrapper around an idempotent local call. Nothing replaces it.\`
-
-✅ \`L30-44: shrink: manual loop builds dict. dict(zip(keys, values)), 1 line.\`
-
-## Scoring
-
-End with the only metric that matters: \`net: -<N> lines possible.\`
-
-If there is nothing to cut, say \`Lean already. Ship.\` and stop.
-
-## Boundaries
-
-Scope: over-engineering and complexity only. Correctness bugs, security holes,
-and performance are explicitly out of scope. Route them to a normal review
-pass, not this one. A single smoke test or \`assert\`-based
-self-check is the ponytail minimum, not bloat, never flag it for deletion.
-Does not apply the fixes, only lists them.
-"stop ponytail-review" or "normal mode": revert to verbose review style.
-`;
-const REVIEW_DESCRIPTION = "Code review focused exclusively on over-engineering. Finds what to delete: reinvented standard library, unneeded dependencies, speculative abstractions, dead flexibility. One line per finding: location, what to cut, what replaces it. Use when the user says \"review for over-engineering\", \"what can we delete\", \"is this over-engineered\", \"simplify review\", or invokes /ponytail-review. Complements correctness-focused review, this one only hunts complexity.";
-const AUDIT_SKILL_BODY = `
-ponytail-review, repo-wide. Scan the whole tree instead of a diff. Rank
-findings biggest cut first.
-
-## Tags
-
-Same as ponytail-review:
-
-- \`delete:\` dead code, unused flexibility, speculative feature. Replacement: nothing.
-- \`stdlib:\` hand-rolled thing the standard library ships. Name the function.
-- \`native:\` dependency or code doing what the platform already does. Name the feature.
-- \`yagni:\` abstraction with one implementation, config nobody sets, layer with one caller.
-- \`shrink:\` same logic, fewer lines. Show the shorter form.
-
-## Hunt
-
-Deps the stdlib or platform already ships, single-implementation interfaces,
-factories with one product, wrappers that only delegate, files exporting one
-thing, dead flags and config, hand-rolled stdlib.
-
-## Output
-
-One line per finding, ranked: \`<tag> <what to cut>. <replacement>. [path]\`.
-End with \`net: -<N> lines, -<M> deps possible.\` Nothing to cut: \`Lean already. Ship.\`
-
-## Boundaries
-
-Scope: over-engineering and complexity only. Correctness bugs, security holes,
-and performance are explicitly out of scope. Route them to a normal review
-pass. Lists findings, applies nothing. One-shot.
-"stop ponytail-audit" or "normal mode" to revert.
-`;
-const AUDIT_DESCRIPTION = "Whole-repo audit for over-engineering. Like ponytail-review, but scans the entire codebase instead of a diff: a ranked list of what to delete, simplify, or replace with stdlib/native equivalents. Use when the user says \"audit this codebase\", \"audit for over-engineering\", \"what can I delete from this repo\", \"find bloat\", \"ponytail-audit\", or /ponytail-audit. One-shot report, does not apply fixes.";
-const DEBT_SKILL_BODY = `
-Every deliberate ponytail shortcut is marked with a \`ponytail:\` comment naming
-its ceiling and upgrade path. This collects them into one ledger so a deferral
-can't quietly become permanent.
-
-## Scan
-
-Grep the repo for comment markers, skipping \`node_modules\`, \`.git\`, and build
-output:
-
-\`grep -rnE '(#|//) ?ponytail:' .\`  (add other comment prefixes if your stack uses them)
+\`<safe-delete|verify-first> <tag> <what>. Replace: <simpler form>. Evidence: <path:line + observed use>.\`
 
-Each hit is one ledger row. The comment prefix keeps prose that merely mentions
-the convention out of the ledger.
-
-## Output
-
-One row per marker, grouped by file:
-
-\`<file>:<line>, <what was simplified>. ceiling: <the limit named>. upgrade: <the trigger to revisit>.\`
-
-The convention is \`ponytail: <ceiling>, <upgrade path>\`, so pull the ceiling
-and the trigger straight from the comment. Want an owner per row too? add
-\`git blame -L<line>,<line>\`.
-
-Flag the rot risk: any \`ponytail:\` comment that names no upgrade path or
-trigger gets a \`no-trigger\` tag, those are the ones that silently rot.
-
-End with \`<N> markers, <M> with no trigger.\` Nothing found: \`No ponytail: debt. Clean ledger.\`
-
-## Boundaries
-
-Reads and reports only, changes nothing. To persist it, ask and it writes the
-ledger to a file (e.g. \`PONYTAIL-DEBT.md\`). One-shot. "stop ponytail-debt" or
-"normal mode" to revert.
-`;
-const DEBT_DESCRIPTION = "Harvest every `ponytail:` comment in the codebase into a debt ledger, so the deliberate shortcuts and deferrals ponytail leaves behind get tracked instead of rotting into \"later means never\". Use when the user says \"ponytail debt\", \"/ponytail-debt\", \"what did ponytail defer\", \"list the shortcuts\", \"ponytail ledger\", or \"what did we mark to do later\". One-shot report, changes nothing.";
-const GAIN_SKILL_BODY = `
-Display this scoreboard when invoked. One-shot: do NOT change mode, write flag
-files, or persist anything.
-
-These are upstream Ponytail results, not measured guarantees for this DSH
-adapter.
-
-Savings depend on model, workload, prompt caching, tool usage, and execution
-path. Already-minimal tasks may show little or no savings. Some reasoning
-models may become more expensive because prompt and reasoning overhead can
-exceed the saved output.
-
-## 1. Upstream agentic reference
-
-Real Claude Code sessions on real repositories; 12 feature tasks:
-
-- Source LOC: ~\u221254%
-- Tokens: ~\u221222%
-- Cost: ~\u221220%
-- Time: ~\u221227%
-- Over-build tasks: \u221260\u201394%
-- Safety tests: 100%
-
-## 2. Upstream single-shot reference
-
-5 everyday tasks (email validator, debounce, CSV sum, countdown timer, rate
-limiter); 3 Claude models; single generation per task:
-
-- Lines of code: \u221280\u201394%
-- Cost (Claude): \u221242\u201375%
-- Latency: ~3.1\u20135.8\u00d7 faster
-
-## 3. DSH adapter status
-
-Current DSH smoke tests provide directional evidence only. Stable token,
-cost, and latency savings have not been established.
-
-See the repository's DSH smoke reports for limited, non-statistical
-directional evidence (docs/dsh-smoke-summary.md).
-
-## 4. Honesty boundary
-
-These are upstream benchmark medians, not this repo and not this DSH
-adapter. NEVER print a per-repo savings number ("you saved X lines/tokens
-here"): the unbuilt version was never written, so there is no real baseline
-to subtract from in a live repo. The only real per-repo figures come from
-\`/ponytail-debt\` (a counted ledger), and this card points there instead of
-inventing one. Never claim "Ponytail always saves tokens/cost" or that this
-adapter reproduces the upstream percentages. A missing cost figure (null) is
-not a zero cost.
-
-## Boundaries
-
-One-shot display. Edits nothing, changes no mode.
-"stop ponytail" or "normal mode": revert.
-`;
-const GAIN_DESCRIPTION = "Less unnecessary work; token, cost, and latency effects depend on model and workload. Upstream benchmark reference, not a DSH-adapter guarantee. One-shot display, not a persistent mode, and not a per-repo number. Trigger: /ponytail-gain, \"ponytail gain\", \"what does ponytail save\", \"show ponytail impact\", \"ponytail scoreboard\".";
-const HELP_SKILL_BODY = `
-Display this reference card when invoked. One-shot, do NOT change mode,
-write flag files, or persist anything.
-
-## Levels
-
-| Level | Trigger | What change |
-|-------|---------|-------------|
-| **Lite** | \`/ponytail lite\` | Build what's asked, name the lazier alternative in one line. |
-| **Full** | \`/ponytail\` | The ladder enforced: YAGNI → stdlib → native → one line → minimum. Default. |
-| **Ultra** | \`/ponytail ultra\` | YAGNI extremist: deletion first, questions speculation — never cuts explicit requirements. |
-| **Off** | \`/ponytail off\` | Ponytail stops injecting its ruleset for this session. |
-
-Level is session-scoped until changed.
-
-## Choosing a level
-
-- **Lite**: Use for small, explicit changes or when the implementation is
-  already clear. Completes explicit requirements without actively
-  challenging them.
-  Lite：小改动、需求明确时使用。
-- **Full**: Use for new features, refactors, root-cause bug fixes, or tasks
-  likely to invite unnecessary abstractions, dependencies, or custom
-  components.
-  Full：新功能、重构、根因修复、容易过度设计时使用。
-- **Ultra**: Use for deliberate code cleanup and over-engineering removal.
-  It questions speculative scope, but never removes explicit requirements,
-  security, validation, accessibility, or data-loss protection.
-  Ultra：专门清理冗余和过度抽象时使用。
-- **Off**: Use when the task is non-coding, already fully specified, or when
-  the fixed prompt overhead is not worthwhile.
-  Off：非编码任务或已经明确到无需额外编码判断的任务。
-
-Ponytail is not a guaranteed token-saving switch. It trades a small fixed
-prompt cost for a chance to reduce unnecessary work. Do not default every
-task to Ultra.
-
-## Skills
-
-| Skill | Trigger | What it does |
-|-------|---------|--------------|
-| **ponytail** | \`/ponytail\` | Lazy mode itself. Simplest solution that works. |
-| **ponytail-review** | \`/ponytail-review\` | Over-engineering review: \`L42: yagni: factory, one product. Inline.\` |
-| **ponytail-audit** | \`/ponytail-audit\` | Whole-repo over-engineering audit: ranked list of what to delete. |
-| **ponytail-debt** | \`/ponytail-debt\` | Harvest \`ponytail:\` shortcut comments into a tracked ledger. |
-| **ponytail-gain** | \`/ponytail-gain\` | Upstream benchmark reference: less unnecessary work; token/cost/latency effects depend on model and workload. |
-| **ponytail-help** | \`/ponytail-help\` | This card. |
-
-You can also load any of these with the \`skill\` tool.
+- \`safe-delete\`: no required consumer or behavior is lost.
+- \`verify-first\`: consumers may remain; name the check needed before deletion.
 
+Tags:
+
+- \`delete:\` dead code or speculative flexibility.
+- \`stdlib:\` hand-rolled standard-library behavior.
+- \`native:\` dependency or code replaced by a platform feature.
+- \`yagni:\` one-use abstraction, unset config, or one-caller layer.
+- \`shrink:\` identical behavior in fewer lines.
+
+Count only concrete ranges and manifest entries. End with
+\`net: -<N> lines, -<M> deps countable.\` Use \`uncounted\` for either unknown.
+Nothing to cut: \`Lean already. Ship.\`
+
+## Boundaries
+
+Scope: over-engineering and complexity only. Correctness bugs, security holes,
+and performance are explicitly out of scope. Route them to a normal review
+pass. Lists findings, applies nothing. One-shot.
+"stop ponytail-audit" or "normal mode" to revert.
+`;
+const AUDIT_DESCRIPTION = "Whole-repo audit for over-engineering. Like ponytail-review, but scans the entire codebase instead of a diff: a ranked list of what to delete, simplify, or replace with stdlib/native equivalents. Use when the user says \"audit this codebase\", \"audit for over-engineering\", \"what can I delete from this repo\", \"find bloat\", \"ponytail-audit\", or /ponytail-audit. One-shot report, does not apply fixes.";
+const DEBT_SKILL_BODY = `
+Collect \`ponytail:\` shortcut comments into a debt ledger. Each should name its
+ceiling and upgrade path.
+
+## Scan
+
+Prefer ripgrep and exclude generated or dependency trees:
+
+\`rg -n --hidden --glob '!node_modules/**' --glob '!.git/**' --glob '!lib/**' --glob '!dist/**' --glob '!build/**' '(#|//) ?ponytail:' .\`
+
+If \`rg\` is unavailable, scan tracked files only:
+\`git grep -n -E '(#|//) ?ponytail:'\`
+
+Each hit is one row; the comment prefix excludes prose mentions.
+
+## Output
+
+One row per marker, grouped by file:
+
+\`<file>:<line>, <what was simplified>. ceiling: <the limit named>. upgrade: <the trigger to revisit>.\`
+
+The convention is \`ponytail: <ceiling>, <upgrade path>\`; copy both from the
+comment. Add an owner only when requested, using
+\`git blame -L<line>,<line> <file>\`.
+
+Flag the rot risk: any \`ponytail:\` comment that names no upgrade path or
+trigger gets a \`no-trigger\` tag, those are the ones that silently rot.
+
+End with \`<N> markers, <M> with no trigger.\` Nothing found: \`No ponytail: debt. Clean ledger.\`
+
+## Boundaries
+
+Reads and reports only, changes nothing. To persist it, ask and it writes the
+ledger to a file (e.g. \`PONYTAIL-DEBT.md\`). One-shot. "stop ponytail-debt" or
+"normal mode" to revert.
+`;
+const DEBT_DESCRIPTION = "Harvest every `ponytail:` comment in the codebase into a debt ledger, so the deliberate shortcuts and deferrals ponytail leaves behind get tracked instead of rotting into \"later means never\". Use when the user says \"ponytail debt\", \"/ponytail-debt\", \"what did ponytail defer\", \"list the shortcuts\", \"ponytail ledger\", or \"what did we mark to do later\". One-shot report, changes nothing.";
+const GAIN_SKILL_BODY = `
+Display this scoreboard when invoked. One-shot: do NOT change mode, write flag
+files, or persist anything.
+
+These are upstream Ponytail results, not measured guarantees for this DSH
+adapter.
+
+Savings depend on model, workload, prompt caching, tool usage, and execution
+path. Already-minimal tasks may show little or no savings. Some reasoning
+models may become more expensive because prompt and reasoning overhead can
+exceed the saved output.
+
+## 1. Upstream agentic reference
+
+Real Claude Code sessions on real repositories; 12 feature tasks:
+
+- Source LOC: ~\u221254%
+- Tokens: ~\u221222%
+- Cost: ~\u221220%
+- Time: ~\u221227%
+- Over-build tasks: \u221260\u201394%
+- Safety tests: 100%
+
+## 2. Upstream single-shot reference
+
+5 everyday tasks (email validator, debounce, CSV sum, countdown timer, rate
+limiter); 3 Claude models; single generation per task:
+
+- Lines of code: \u221280\u201394%
+- Cost (Claude): \u221242\u201375%
+- Latency: ~3.1\u20135.8\u00d7 faster
+
+## 3. DSH adapter status
+
+Current DSH smoke tests provide directional evidence only. Stable token,
+cost, and latency savings have not been established.
+
+See the repository's DSH smoke reports for limited, non-statistical
+directional evidence (docs/dsh-smoke-summary.md).
+
+## 4. Honesty boundary
+
+These are upstream benchmark medians, not this repo and not this DSH
+adapter. NEVER print a per-repo savings number ("you saved X lines/tokens
+here"): the unbuilt version was never written, so there is no real baseline
+to subtract from in a live repo. The only real per-repo figures come from
+\`/ponytail-debt\` (a counted ledger), and this card points there instead of
+inventing one. Never claim "Ponytail always saves tokens/cost" or that this
+adapter reproduces the upstream percentages. A missing cost figure (null) is
+not a zero cost.
+
+## Boundaries
+
+One-shot display. Edits nothing, changes no mode.
+"stop ponytail" or "normal mode": revert.
+`;
+const GAIN_DESCRIPTION = "Less unnecessary work; token, cost, and latency effects depend on model and workload. Upstream benchmark reference, not a DSH-adapter guarantee. One-shot display, not a persistent mode, and not a per-repo number. Trigger: /ponytail-gain, \"ponytail gain\", \"what does ponytail save\", \"show ponytail impact\", \"ponytail scoreboard\".";
+const HELP_SKILL_BODY = `
+Display this one-shot reference card. Do not change mode or persist anything.
+
+## Levels
+
+| Level | Trigger | What change |
+|-------|---------|-------------|
+| **Lite** | \`/ponytail lite\` | Build what's asked, name the lazier alternative in one line. |
+| **Full** | \`/ponytail\` | The ladder enforced: YAGNI → stdlib → native → one line → minimum. Default. |
+| **Ultra** | \`/ponytail ultra\` | YAGNI extremist: deletion first, questions speculation — never cuts explicit requirements. |
+| **Off** | \`/ponytail off\` | Ponytail stops injecting its ruleset for this session. |
+
+Level is session-scoped until changed.
+
+## Choosing a level
+
+- **Lite**: Use for small, explicit changes or when the implementation is
+  already clear. Completes explicit requirements without actively
+  challenging them.
+  Lite：小改动、需求明确时使用。
+- **Full**: Use for new features, refactors, root-cause bug fixes, or tasks
+  likely to invite unnecessary abstractions, dependencies, or custom
+  components.
+  Full：新功能、重构、根因修复、容易过度设计时使用。
+- **Ultra**: Use for deliberate code cleanup and over-engineering removal.
+  It questions speculative scope, but never removes explicit requirements,
+  security, validation, accessibility, or data-loss protection.
+  Ultra：专门清理冗余和过度抽象时使用。
+- **Off**: Use when the task is non-coding, already fully specified, or when
+  the fixed prompt overhead is not worthwhile.
+  Off：非编码任务或已经明确到无需额外编码判断的任务。
+
+Ponytail is not a guaranteed token-saving switch. It trades a small fixed
+prompt cost for a chance to reduce unnecessary work. Do not default every
+task to Ultra.
+
+## Skills
+
+| Skill | Trigger | What it does |
+|-------|---------|--------------|
+| **ponytail** | \`/ponytail\` | Lazy mode itself. Simplest solution that works. |
+| **ponytail-review** | \`/ponytail-review\` | Over-engineering review: \`L42: yagni: factory, one product. Inline.\` |
+| **ponytail-audit** | \`/ponytail-audit\` | Whole-repo over-engineering audit: ranked list of what to delete. |
+| **ponytail-debt** | \`/ponytail-debt\` | Harvest \`ponytail:\` shortcut comments into a tracked ledger. |
+| **ponytail-gain** | \`/ponytail-gain\` | Upstream benchmark reference: less unnecessary work; token/cost/latency effects depend on model and workload. |
+| **ponytail-help** | \`/ponytail-help\` | This card. |
+
 ## Deactivate
 
-Say "stop ponytail" or "normal mode". Resume anytime with \`/ponytail\` —
-it re-enables at the effective default (or \`full\` when that is off too).
-\`/ponytail status\` only shows the current level, never changes it.
-\`/ponytail off\` also works. Level is session-scoped; a new session starts
-from the configured default.
-
-## Configure Default Mode
-
-Default mode = \`full\`, auto-active every session. Change it:
-
-**Environment variable** (highest priority):
-\`\`\`bash
-export PONYTAIL_DEFAULT_MODE=ultra
-\`\`\`
-
-**Config file** (\`~/.config/ponytail/config.json\`, Windows: \`%APPDATA%\\ponytail\\config.json\`):
-\`\`\`json
-{ "defaultMode": "lite" }
-\`\`\`
-
-**Profile config** (per DSH profile, via the bundle row's \`config\` — e.g.
-\`tui\` → lite):
-
-\`\`\`yaml
-- insert:
-    - id: ponytail
-      name: '@mengyuly/dsh-ponytail'
-      config:
-        defaultMode: lite
-\`\`\`
-
-Set \`"off"\` to disable auto-activation on session start, activate manually
-with \`/ponytail\` when wanted. \`/ponytail default <mode>\` persists a new
-default to the user config file; an exported \`PONYTAIL_DEFAULT_MODE\` or a
-profile \`defaultMode\` still outranks the saved value for new sessions.
-
-Resolution: session override > env var > profile config > config file > \`full\`.
-
-## More
-
-Full docs + examples: https://github.com/DietrichGebert/ponytail
-`;
-const HELP_DESCRIPTION = "Quick-reference card for all ponytail modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /ponytail-help, \"ponytail help\", \"what ponytail commands\", \"how do I use ponytail\".";
-/** Ordered set of runtime skills surfaced to the model catalog and `/` menu. */
-function ponytailSkills() {
-	return [
-		{
-			name: "ponytail",
-			source: "runtime",
-			description: PONYTAIL_DESCRIPTION,
-			whenToUse: "Use only when the user asks about Ponytail activation, modes, configuration, or help. Coding tasks already receive the active ruleset from the system prompt.",
-			content: PONYTAIL_SKILL_BODY,
-			invocation: {
-				modelInvocable: false,
-				userInvocable: true
-			}
-		},
-		{
-			name: "ponytail-review",
-			source: "runtime",
-			description: REVIEW_DESCRIPTION,
-			content: REVIEW_SKILL_BODY,
-			invocation: {
-				modelInvocable: true,
-				userInvocable: true
-			}
-		},
-		{
-			name: "ponytail-audit",
-			source: "runtime",
-			description: AUDIT_DESCRIPTION,
-			content: AUDIT_SKILL_BODY,
-			invocation: {
-				modelInvocable: true,
-				userInvocable: true
-			}
-		},
-		{
-			name: "ponytail-debt",
-			source: "runtime",
-			description: DEBT_DESCRIPTION,
-			content: DEBT_SKILL_BODY,
-			invocation: {
-				modelInvocable: true,
-				userInvocable: true
-			}
-		},
-		{
-			name: "ponytail-gain",
-			source: "runtime",
-			description: GAIN_DESCRIPTION,
-			content: GAIN_SKILL_BODY,
-			invocation: {
-				modelInvocable: true,
-				userInvocable: true
-			}
-		},
-		{
-			name: "ponytail-help",
-			source: "runtime",
-			description: HELP_DESCRIPTION,
-			content: HELP_SKILL_BODY,
-			invocation: {
-				modelInvocable: true,
-				userInvocable: true
-			}
-		}
-	];
-}
-//#endregion
-//#region lib/types/modes.js
-/**
-* Ponytail mode resolution: the effective default comes from, in order, the
-* `PONYTAIL_DEFAULT_MODE` environment variable, the Cordis profile
-* `defaultMode`, the optional user config file
-* `~/.config/ponytail/config.json` (`defaultMode`), then `full`. Setting a
-* level via the `/ponytail` command is session-scoped and lives in an
-* in-memory, per-agent {@link ModeStore}.
-*
-* @module @deepseek-ai/dsh-ponytail
-*/
-const DEFAULT_MODE = "full";
-const RUNTIME_MODES = [
-	"off",
-	"lite",
-	"full",
-	"ultra"
-];
-/** Strip a UTF-8 BOM that Windows editors prepend before JSON.parse. */
-function stripBom(text) {
-	return text.replace(/^\uFEFF/, "");
-}
-/**
-* Normalize free-form input to a runtime intensity. `null` for anything that
-* is not exactly `off`, `lite`, `full`, or `ultra`.
-*/
-function normalizeRuntimeMode(mode) {
-	if (typeof mode !== "string") return null;
-	const normalized = mode.trim().toLowerCase();
-	return RUNTIME_MODES.includes(normalized) ? normalized : null;
-}
-/**
-* Deactivation commands only match when the whole message is the command,
-* ignoring case and trailing punctuation. Matching the phrase anywhere would
-* turn ponytail off mid-task for ordinary requests like "add a normal mode
-* toggle".
-*/
-function isDeactivationCommand(text) {
-	const normalized = (typeof text === "string" ? text : "").trim().toLowerCase().replace(/[\s.!?。？！]+$/, "");
-	return normalized === "stop ponytail" || normalized === "normal mode";
-}
-/** Config directory: `$XDG_CONFIG_HOME/ponytail`, `%APPDATA%\ponytail`, else `~/.config/ponytail`. */
-function configDir(env = process.env) {
-	if (env.XDG_CONFIG_HOME) return join(env.XDG_CONFIG_HOME, "ponytail");
-	if (process.platform === "win32") return join(env.APPDATA || join(homedir(), "AppData", "Roaming"), "ponytail");
-	return join(homedir(), ".config", "ponytail");
-}
-/** Absolute path of the optional `config.json`. */
-function configPath(env = process.env) {
-	return join(configDir(env), "config.json");
-}
-/**
-* Read the configured default with diagnostics. Priority:
-* `PONYTAIL_DEFAULT_MODE` → Cordis profile `defaultMode` → user config file →
-* `full`. A missing config file is normal and yields no issue; a broken one
-* yields the fallback mode plus one issue for the caller to warn about once.
-* @param env - the process environment to read.
-* @param profileMode - the validated Cordis profile `defaultMode`, or `null`
-*   when the profile config is absent or invalid (invalid values are reported
-*   by the caller; this function only consumes valid ones).
-*/
-function readDefaultModeInfo(env = process.env, profileMode = null) {
-	const path = configPath(env);
-	const envMode = normalizeRuntimeMode(env.PONYTAIL_DEFAULT_MODE);
-	let configText;
-	try {
-		configText = readFileSync(path, "utf8");
-	} catch (error) {
-		if (error.code === "ENOENT") return { mode: envMode ?? profileMode ?? "full" };
-		return {
-			mode: envMode ?? profileMode ?? "full",
-			issue: {
-				kind: "read",
-				detail: `${path}: ${error.message}`
-			}
-		};
-	}
-	let configIssue;
-	let fromConfig = null;
-	try {
-		const parsed = JSON.parse(stripBom(configText));
-		if (!parsed || typeof parsed !== "object" || Array.isArray(parsed)) configIssue = {
-			kind: "shape",
-			detail: `${path}: root must be a JSON object`
-		};
-		else if ("defaultMode" in parsed) {
-			fromConfig = normalizeRuntimeMode(parsed.defaultMode);
-			if (!fromConfig) configIssue = {
-				kind: "value",
-				detail: `${path}: defaultMode is not lite|full|ultra|off`
-			};
-		}
-	} catch (error) {
-		configIssue = {
-			kind: "json",
-			detail: `${path}: ${error.message}`
-		};
-	}
-	if (envMode) return {
-		mode: envMode,
-		...configIssue ? { issue: configIssue } : {}
-	};
-	if (profileMode) return {
-		mode: profileMode,
-		...configIssue ? { issue: configIssue } : {}
-	};
-	if (configIssue) return {
-		mode: DEFAULT_MODE,
-		issue: configIssue
-	};
-	return { mode: fromConfig ?? "full" };
-}
-/**
-* Read the configured default for this host: environment variable first, then
-* the Cordis profile `defaultMode`, then the user config file, then `full`.
-*/
-function readDefaultMode(env = process.env, profileMode = null) {
-	return readDefaultModeInfo(env, profileMode).mode;
-}
-/**
-* Why a `saved` default is not the effective one — for the `/ponytail default`
-* result message. `null` means the saved value is effective.
-*/
-function defaultOverrideReason(env, profileMode) {
-	if (normalizeRuntimeMode(env.PONYTAIL_DEFAULT_MODE)) return "PONYTAIL_DEFAULT_MODE";
-	if (profileMode) return "profile configuration";
-	return null;
-}
-/**
-* Persist a new default level to the config file, preserving other fields.
-* Returns the normalized mode, or `null` when the value is not a runtime mode.
-* Throws when the write itself fails, so callers never report success for a
-* file that was not written.
-*/
-function writeDefaultMode(mode, env = process.env) {
-	const normalized = normalizeRuntimeMode(mode);
-	if (!normalized) return null;
-	const path = configPath(env);
-	let config = {};
-	try {
-		const parsed = JSON.parse(stripBom(readFileSync(path, "utf8")));
-		if (parsed && typeof parsed === "object" && !Array.isArray(parsed)) config = parsed;
-	} catch {}
-	config.defaultMode = normalized;
-	const text = `${JSON.stringify(config, null, 2)}\n`;
-	mkdirSync(dirname(path), { recursive: true });
-	const temp = join(dirname(path), `.config-${process.pid}-${Date.now()}.tmp`);
-	try {
-		writeFileSync(temp, text, "utf8");
-		renameSync(temp, path);
-	} catch (error) {
-		try {
-			unlinkSync(temp);
-		} catch {}
-		throw new Error(`failed to write ${path}: ${error.message}`);
-	}
-	return normalized;
-}
-/**
-* Session-scoped live mode. The absence of an entry means "use the configured
-* default", which matches the upstream behavior where each session starts from
-* the default until the user switches it.
-*/
-var ModeStore = class {
-	modes = /* @__PURE__ */ new Map();
-	/** The mode in force for one agent, or the configured default. */
+Say "stop ponytail", "normal mode", "停止 ponytail", or "正常模式".
+\`/ponytail\` re-enables at the effective default (or \`full\` if it is \`off\`).
+\`/ponytail reset\` clears the session override and follows the configured default.
+\`/ponytail status\` reports mode and source; \`/ponytail off\` disables it.
+
+## Configure Default Mode
+
+The built-in fallback is \`full\`; check the effective mode with
+\`/ponytail status\`. Change the default:
+
+**Environment variable** (highest priority):
+\`\`\`bash
+export PONYTAIL_DEFAULT_MODE=ultra
+\`\`\`
+
+**Config file** (\`~/.config/ponytail/config.json\`, Windows: \`%APPDATA%\\ponytail\\config.json\`):
+\`\`\`json
+{ "defaultMode": "lite" }
+\`\`\`
+
+**Profile config** (per DSH profile, via the bundle row's \`config\` — e.g.
+\`tui\` → lite):
+
+\`\`\`yaml
+- insert:
+    - id: ponytail
+      name: '@mengyuly/dsh-ponytail'
+      config:
+        defaultMode: lite
+\`\`\`
+
+Set \`"off"\` to disable auto-activation on session start, activate manually
+with \`/ponytail\` when wanted. \`/ponytail default <mode>\` persists a new
+default to the user config file; an exported \`PONYTAIL_DEFAULT_MODE\` or a
+profile \`defaultMode\` still outranks the saved value for new sessions.
+
+Resolution: session override > env var > profile config > config file > \`full\`.
+
+## More
+
+Full docs + examples: https://github.com/DietrichGebert/ponytail
+`;
+const HELP_DESCRIPTION = "Quick-reference card for all ponytail modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /ponytail-help, \"ponytail help\", \"what ponytail commands\", \"how do I use ponytail\".";
+/** Ordered set of runtime skills surfaced to the model catalog and `/` menu. */
+function ponytailSkills() {
+	return [
+		{
+			name: "ponytail",
+			source: "runtime",
+			description: PONYTAIL_DESCRIPTION,
+			whenToUse: "Use only when the user asks about Ponytail activation, modes, configuration, or help. Coding tasks already receive the active ruleset from the system prompt.",
+			content: PONYTAIL_SKILL_BODY,
+			invocation: {
+				modelInvocable: false,
+				userInvocable: true
+			}
+		},
+		{
+			name: "ponytail-review",
+			source: "runtime",
+			description: REVIEW_DESCRIPTION,
+			content: REVIEW_SKILL_BODY,
+			invocation: {
+				modelInvocable: true,
+				userInvocable: true
+			}
+		},
+		{
+			name: "ponytail-audit",
+			source: "runtime",
+			description: AUDIT_DESCRIPTION,
+			content: AUDIT_SKILL_BODY,
+			invocation: {
+				modelInvocable: true,
+				userInvocable: true
+			}
+		},
+		{
+			name: "ponytail-debt",
+			source: "runtime",
+			description: DEBT_DESCRIPTION,
+			content: DEBT_SKILL_BODY,
+			invocation: {
+				modelInvocable: true,
+				userInvocable: true
+			}
+		},
+		{
+			name: "ponytail-gain",
+			source: "runtime",
+			description: GAIN_DESCRIPTION,
+			content: GAIN_SKILL_BODY,
+			invocation: {
+				modelInvocable: true,
+				userInvocable: true
+			}
+		},
+		{
+			name: "ponytail-help",
+			source: "runtime",
+			description: HELP_DESCRIPTION,
+			content: HELP_SKILL_BODY,
+			invocation: {
+				modelInvocable: true,
+				userInvocable: true
+			}
+		}
+	];
+}
+//#endregion
+//#region lib/types/modes.js
+/**
+* Ponytail mode resolution: the effective default comes from, in order, the
+* `PONYTAIL_DEFAULT_MODE` environment variable, the Cordis profile
+* `defaultMode`, the optional user config file
+* `~/.config/ponytail/config.json` (`defaultMode`), then `full`. Setting a
+* level via the `/ponytail` command is session-scoped and lives in an
+* in-memory, per-agent {@link ModeStore}.
+*
+* @module @mengyuly/dsh-ponytail
+*/
+const DEFAULT_MODE = "full";
+const RUNTIME_MODES = [
+	"off",
+	"lite",
+	"full",
+	"ultra"
+];
+/** Strip a UTF-8 BOM that Windows editors prepend before JSON.parse. */
+function stripBom(text) {
+	return text.replace(/^\uFEFF/, "");
+}
+/**
+* Normalize free-form input to a runtime intensity. `null` for anything that
+* is not exactly `off`, `lite`, `full`, or `ultra`.
+*/
+function normalizeRuntimeMode(mode) {
+	if (typeof mode !== "string") return null;
+	const normalized = mode.trim().toLowerCase();
+	return RUNTIME_MODES.includes(normalized) ? normalized : null;
+}
+/**
+* English and Chinese deactivation commands match only when the whole message is the command,
+* ignoring case and trailing punctuation. Matching the phrase anywhere would
+* turn ponytail off mid-task for ordinary requests like "add a normal mode
+* toggle".
+*/
+function isDeactivationCommand(text) {
+	const normalized = (typeof text === "string" ? text : "").trim().toLowerCase().replace(/[\s.!?。？！]+$/, "");
+	return normalized === "stop ponytail" || normalized === "normal mode" || normalized === "停止 ponytail" || normalized === "关闭 ponytail" || normalized === "普通模式" || normalized === "正常模式";
+}
+/** Config directory: `$XDG_CONFIG_HOME/ponytail`, `%APPDATA%\ponytail`, else `~/.config/ponytail`. */
+function configDir(env = process.env) {
+	if (env.XDG_CONFIG_HOME) return join(env.XDG_CONFIG_HOME, "ponytail");
+	if (process.platform === "win32") return join(env.APPDATA || join(homedir(), "AppData", "Roaming"), "ponytail");
+	return join(homedir(), ".config", "ponytail");
+}
+/** Absolute path of the optional `config.json`. */
+function configPath(env = process.env) {
+	return join(configDir(env), "config.json");
+}
+/**
+* Read the configured default with diagnostics. Priority:
+* `PONYTAIL_DEFAULT_MODE` → Cordis profile `defaultMode` → user config file →
+* `full`. A missing config file is normal and yields no issue; a broken one
+* yields the fallback mode plus one issue for the caller to warn about once.
+* @param env - the process environment to read.
+* @param profileMode - the validated Cordis profile `defaultMode`, or `null`
+*   when the profile config is absent or invalid (invalid values are reported
+*   by the caller; this function only consumes valid ones).
+*/
+function readDefaultModeInfo(env = process.env, profileMode = null) {
+	const path = configPath(env);
+	const envMode = normalizeRuntimeMode(env.PONYTAIL_DEFAULT_MODE);
+	let configText;
+	try {
+		configText = readFileSync(path, "utf8");
+	} catch (error) {
+		if (error.code === "ENOENT") return { mode: envMode ?? profileMode ?? "full" };
+		return {
+			mode: envMode ?? profileMode ?? "full",
+			issue: {
+				kind: "read",
+				detail: `${path}: ${error.message}`
+			}
+		};
+	}
+	let configIssue;
+	let fromConfig = null;
+	try {
+		const parsed = JSON.parse(stripBom(configText));
+		if (!parsed || typeof parsed !== "object" || Array.isArray(parsed)) configIssue = {
+			kind: "shape",
+			detail: `${path}: root must be a JSON object`
+		};
+		else if ("defaultMode" in parsed) {
+			fromConfig = normalizeRuntimeMode(parsed.defaultMode);
+			if (!fromConfig) configIssue = {
+				kind: "value",
+				detail: `${path}: defaultMode is not lite|full|ultra|off`
+			};
+		}
+	} catch (error) {
+		configIssue = {
+			kind: "json",
+			detail: `${path}: ${error.message}`
+		};
+	}
+	if (envMode) return {
+		mode: envMode,
+		...configIssue ? { issue: configIssue } : {}
+	};
+	if (profileMode) return {
+		mode: profileMode,
+		...configIssue ? { issue: configIssue } : {}
+	};
+	if (configIssue) return {
+		mode: DEFAULT_MODE,
+		issue: configIssue
+	};
+	return { mode: fromConfig ?? "full" };
+}
+/**
+* Read the configured default for this host: environment variable first, then
+* the Cordis profile `defaultMode`, then the user config file, then `full`.
+*/
+function readDefaultMode(env = process.env, profileMode = null) {
+	return readDefaultModeInfo(env, profileMode).mode;
+}
+/**
+* Why a `saved` default is not the effective one — for the `/ponytail default`
+* result message. `null` means the saved value is effective.
+*/
+function defaultOverrideReason(env, profileMode) {
+	if (normalizeRuntimeMode(env.PONYTAIL_DEFAULT_MODE)) return "PONYTAIL_DEFAULT_MODE";
+	if (profileMode) return "profile configuration";
+	return null;
+}
+/**
+* Persist a new default level to the config file, preserving other fields.
+* Returns the normalized mode, or `null` when the value is not a runtime mode.
+* Throws when the write itself fails, so callers never report success for a
+* file that was not written.
+*/
+function writeDefaultMode(mode, env = process.env) {
+	const normalized = normalizeRuntimeMode(mode);
+	if (!normalized) return null;
+	const path = configPath(env);
+	let config = {};
+	try {
+		const parsed = JSON.parse(stripBom(readFileSync(path, "utf8")));
+		if (parsed && typeof parsed === "object" && !Array.isArray(parsed)) config = parsed;
+	} catch {}
+	config.defaultMode = normalized;
+	const text = `${JSON.stringify(config, null, 2)}\n`;
+	mkdirSync(dirname(path), { recursive: true });
+	const temp = join(dirname(path), `.config-${process.pid}-${Date.now()}.tmp`);
+	try {
+		writeFileSync(temp, text, "utf8");
+		renameSync(temp, path);
+	} catch (error) {
+		try {
+			unlinkSync(temp);
+		} catch {}
+		throw new Error(`failed to write ${path}: ${error.message}`);
+	}
+	return normalized;
+}
+/**
+* Session-scoped live mode. The absence of an entry means "use the configured
+* default", which matches the upstream behavior where each session starts from
+* the default until the user switches it.
+*/
+var ModeStore = class {
+	modes = /* @__PURE__ */ new Map();
+	/** The mode in force for one agent, or the configured default. */
 	modeFor(agentId, fallback) {
 		return this.modes.get(agentId) ?? fallback;
 	}
-	/** Set the mode for one agent's session (session-scoped, survives until changed or disposal). */
-	set(agentId, mode) {
-		this.modes.set(agentId, mode);
-	}
-	/** Forget a session-scoped override so the next lookup returns the default. */
-	clear(agentId) {
-		this.modes.delete(agentId);
-	}
-};
-/**
-* Compile `PONYTAIL_SUBAGENT_MATCHER` into a case-insensitive regex. An unset
-* matcher yields `null`; an invalid pattern stays fail-open (every agent gets
-* the ruleset) but is reported so the caller can warn exactly once.
-*/
-function compileSubagentMatcher(raw) {
-	if (!raw) return {
-		matcher: null,
-		invalid: false
-	};
-	try {
-		return {
-			matcher: new RegExp(raw, "i"),
-			invalid: false
-		};
-	} catch {
-		return {
-			matcher: null,
-			invalid: true
-		};
+	/** Whether this session currently overrides the configured default. */
+	has(agentId) {
+		return this.modes.has(agentId);
 	}
-}
-/**
-* The stable per-session identity backing every mode override. DSH's `Agent`
-* type documents `id` as "the single identity shared with session", so the
-* agent id IS the SessionId: one entry per live session, never shared between
-* two sessions, and stable across the session's lifetime. Centralized so the
-* key choice lives in exactly one place.
-*/
-function sessionKey(agent) {
-	return agent.id;
-}
-/** Whether a session is a subagent child (origin, or any delegation depth with no origin). */
-function isSubagentSession(header) {
-	return header.origin === "subagent" || (header.delegationDepth ?? 0) > 0;
-}
-//#endregion
-//#region lib/types/instructions.js
-/**
-* Structured ponytail ruleset composition. Each intensity is built from
-* explicit fragments — common rules, a never-cut safety boundary list, and
-* the mode's own rules — instead of filtering one Markdown body with regexes.
-* The three intensities therefore differ in their actual instructions, not
-* just in a table row.
-*
-* @module @deepseek-ai/dsh-ponytail
-*/
-/** Shared identity line, carried by every non-`off` mode. */
-const INTRO = "You are a lazy senior developer. Lazy means efficient, not careless. The best code is the code never written.";
-/**
-* Understanding-and-reuse baseline, identical in every non-`off` mode.
-*/
+	/** Set the mode for one agent's session (session-scoped, survives until changed or disposal). */
+	set(agentId, mode) {
+		this.modes.set(agentId, mode);
+	}
+	/** Forget a session-scoped override so the next lookup returns the default. */
+	clear(agentId) {
+		this.modes.delete(agentId);
+	}
+};
+/**
+* Compile `PONYTAIL_SUBAGENT_MATCHER` into a case-insensitive regex. An unset
+* matcher yields `null`; an invalid pattern stays fail-open (every agent gets
+* the ruleset) but is reported so the caller can warn exactly once.
+*/
+function compileSubagentMatcher(raw) {
+	if (!raw) return {
+		matcher: null,
+		invalid: false
+	};
+	try {
+		return {
+			matcher: new RegExp(raw, "i"),
+			invalid: false
+		};
+	} catch {
+		return {
+			matcher: null,
+			invalid: true
+		};
+	}
+}
+/**
+* The stable per-session identity backing every mode override. DSH's `Agent`
+* type documents `id` as "the single identity shared with session", so the
+* agent id IS the SessionId: one entry per live session, never shared between
+* two sessions, and stable across the session's lifetime. Centralized so the
+* key choice lives in exactly one place.
+*/
+function sessionKey(agent) {
+	return agent.id;
+}
+/** Whether a session is a subagent child (origin, or any delegation depth with no origin). */
+function isSubagentSession(header) {
+	return header.origin === "subagent" || (header.delegationDepth ?? 0) > 0;
+}
+//#endregion
+//#region lib/types/instructions.js
+/**
+* Structured ponytail ruleset composition. Each intensity is built from
+* explicit fragments — common rules, a never-cut safety boundary list, and
+* the mode's own rules — instead of filtering one Markdown body with regexes.
+* The three intensities therefore differ in their actual instructions, not
+* just in a table row.
+*
+* @module @mengyuly/dsh-ponytail
+*/
+/** Shared identity line, carried by every non-`off` mode. */
+const INTRO = "You are a lazy senior developer. Lazy means efficient, not careless. The best code is the code never written.";
+/**
+* Understanding-and-reuse baseline, identical in every non-`off` mode.
+*/
 const COMMON_RULES = [
-	"Understand the problem before choosing a solution: read the code the change touches and trace the real flow end to end. Laziness that skips comprehension ships a confident wrong fix.",
-	"Reuse what already exists in this codebase before writing anything new.",
-	"Reach for the standard library, platform-native features, and already-installed dependencies before custom code.",
-	"A non-trivial change leaves ONE minimal runnable check behind (an assert-based self-check or one small test file; no frameworks). Trivial one-liners need no test.",
-	"Explain briefly, but never omit the key decisions."
-].join("\n");
-/**
-* The never-cut list. Every non-`off` mode keeps these; intensities tune how
-* aggressively code is minimized, never what may be dropped.
-*/
-const SAFETY_BOUNDARIES = [
-	"Never cut, in any mode:",
-	"- Input validation at trust boundaries.",
-	"- Error handling that prevents data loss.",
-	"- Security measures.",
-	"- Accessibility basics.",
-	"- Explicit acceptance criteria the user asked for.",
-	"- Understanding the problem and tracing the real flow first.",
-	"- The real end-to-end data flow: no UI-only field, unused state, placeholder path, or disconnected payload.",
-	"- Necessary tests for non-trivial changes.",
-	"- Root-cause fixes over symptom patches.",
-	"- \"Minimal diff\" is not a substitute for \"correct fix\"."
-].join("\n");
-/** Lite: complete the explicit ask; reuse; suggest, do not challenge. */
-const LITE_RULES = [
-	"Complete everything explicitly requested, including every acceptance criterion.",
-	"Prefer existing code, standard-library features, native platform features, and already-installed dependencies.",
-	"You may mention a simpler alternative briefly, but do not challenge or reject an explicit requirement.",
-	"Do not change the existing architecture merely to reduce line count.",
-	"Keep the smallest reasonable validation for non-trivial changes."
+	"Before editing, define a concrete, observable done condition; preserve explicit acceptance criteria.",
+	"Read touched code and trace the relevant flow. The ladder is a reflex, not a research project: investigate only what can change the solution, then stop.",
+	"Resolve uncertainty with evidence from code, tools, or authoritative docs; never invent facts or checks.",
+	"Reuse existing code, native features, or installed dependencies before custom code.",
+	"Choose the smallest complete change compatible with existing contracts, not the smallest local diff.",
+	"Loop: inspect, change, run the narrowest relevant check, inspect the final diff. On failure, fix the cause. Do not weaken a test to pass.",
+	"Report only verified results, checks run, and uncertainty; be brief."
 ].join("\n");
-/** Smallest complete end-to-end change: shared by Full and Ultra. */
-const E2E_RULES = [
-	"Smallest complete end-to-end change:",
-	"- Prefer the smallest complete end-to-end change compatible with the existing architecture, not merely the fewest lines in one file.",
-	"- Before creating a component, abstraction, protocol, migration, transport format, storage format, or dependency, inspect the repository’s existing path and preserve its current contract.",
-	"- Do not redesign transport, storage, API shape, or persistence when the task only asks for a local UI or behavior change.",
-	"- A locally smaller implementation that changes the system contract is not smaller overall.",
-	"- Prefer the smallest complete change across the real data flow: input → state → validation → payload → API → persistence → response/UI.",
-	"- It does not mean every layer must change: the change must be complete across the layers it touches.",
-	"- Do not leave a UI-only field, unused state, placeholder path, or disconnected payload merely because it produces a smaller diff."
-].join("\n");
-const MODE_RULES = {
-	lite: LITE_RULES,
-	full: [
-		"Use the complete ladder — stop at the first rung that holds:",
-		"1. Does this need to exist at all? (YAGNI)",
-		"2. Does it already exist in this codebase? Reuse it.",
-		"3. Does the standard library do it? Use it.",
-		"4. Does a native platform feature cover it? Use it.",
-		"5. Does an already-installed dependency solve it? Use it.",
-		"6. Can the solution be reduced to a small expression? Make it that small.",
-		"7. Only then: write the minimum new implementation.",
-		"Default to the shortest correct implementation; prefer deletion and reuse, but do not trade away correctness, security, tests, explicit requirements, or the existing system contract.",
-		"Fix root causes, not symptoms: one guard in the shared function beats a guard in every caller.",
-		"",
-		E2E_RULES
-	].join("\n"),
-	ultra: [
-		"Delete before adding.",
-		"Actively question speculative features, caches, abstractions, configuration, migrations, transport changes, storage changes, and new dependencies.",
-		"Prefer the smallest complete end-to-end change, not the smallest local diff.",
-		"Do not change an existing contract merely to reduce lines.",
-		"For complex requests, ship the smallest correct complete version and state what would justify a larger version.",
-		"Ultra is not refusal: explicit requirements, safety, validation, accessibility, data protection, and acceptance criteria remain mandatory.",
-		"",
-		E2E_RULES
-	].join("\n")
-};
-const MODE_LABELS = {
-	lite: "Lite",
-	full: "Full",
-	ultra: "Ultra"
-};
-/** Compose the complete section text for one intensity. */
-function render(effective) {
-	return [
-		`PONYTAIL MODE ACTIVE — level: ${effective}`,
-		"",
-		INTRO,
-		"",
-		"## Common rules (all modes)",
-		COMMON_RULES,
-		"",
-		"## Safety boundaries (never cut)",
-		SAFETY_BOUNDARIES,
-		"",
-		`## ${MODE_LABELS[effective]} rules`,
-		MODE_RULES[effective]
-	].join("\n");
-}
-/**
-* The injected ruleset for one intensity, composed from the structured
-* fragments above. Returns an empty string for `off` (ponytail contributes
-* nothing). Renders are pure per mode and cached so every turn's bytes stay
-* identical.
-*/
-function getPonytailInstructions(mode) {
-	const effective = normalizeRuntimeMode(mode) ?? "full";
-	if (effective === "off") return "";
-	const cached = instructionCache.get(effective);
-	if (cached !== void 0) return cached;
-	const rendered = render(effective);
-	instructionCache.set(effective, rendered);
-	return rendered;
-}
-/** Rendered rulesets are pure per mode; cache to keep every turn's bytes identical. */
-const instructionCache = /* @__PURE__ */ new Map();
-//#endregion
-//#region lib/types/index.js
-/**
-* Ponytail: the "lazy senior developer" persona as a DeepSeek Harness plugin.
-*
-* One system-prompt section injects the mode-filtered ruleset every turn (the
-* always-on adapter), six runtime skills surface the review/audit/debt/gain/
-* help one-shots, six slash commands drive them from the command plane, and an
-* `agent/pre-step` listener honors the plain-text deactivation phrases.
-*
-* Mode is session-scoped and held in memory; the configured default resolves
-* from `PONYTAIL_DEFAULT_MODE`, then the Cordis profile `defaultMode`, then
-* `~/.config/ponytail/config.json` (see {@link readDefaultMode}), then
-* `full`. A session override via `/ponytail` outranks all of them.
-*
-* @module @deepseek-ai/dsh-ponytail
-*/
-const name = "ponytail";
-const inject = ["systemPrompt", "skills"];
-/** Prompt-section order: after the deployment persona (0), before tool guidance (100–199). */
-const SECTION_ORDER = 40;
-/** Build the one text-line notification a mode switch leaves for the model. */
-function modeNotice(mode) {
-	return mode === "off" ? "PONYTAIL MODE OFF" : `PONYTAIL MODE CHANGED — level: ${mode}`;
-}
-/** Extract the plain text of one user message (only its text blocks). */
-function messageText(message) {
-	const parts = [];
-	for (const block of message.content) if (block.type === "text" && typeof block.text === "string") parts.push(block.text);
-	return parts.join("\n");
-}
-/** Whether any message in a claimed batch is exactly a deactivation command. */
-function containsDeactivation(messages) {
-	return messages.some((message) => isDeactivationCommand(messageText(message)));
-}
-/** Mode visible to one agent: its session override, else the configured default. */
-function modeFor(deps, agent) {
-	return deps.store.modeFor(sessionKey(agent), deps.defaultMode());
-}
-/**
-* Queue one skill's full `<skill_content>` rendering as the model's next
-* ordinary turn, with the same user-explicit `skill-invocation` source the
-* built-in gesture boundary uses.
-*/
-async function queueSkill(deps, invocation, skill) {
-	const loaded = await deps.ctx.skills.get(skill, {
-		cwd: invocation.agent.session.header.cwd,
-		signal: invocation.signal
-	});
-	if (loaded === void 0) return {
-		kind: "error",
-		text: `skill "${skill}" is not available`
-	};
-	const notes = invocation.rawInput.trim();
-	const text = renderSkillContent(loaded) + (notes === "" ? "" : `\n\n${notes}`);
-	invocation.agent.followup(createUserMessage({
-		content: [{
-			type: "text",
-			text
-		}],
-		source: {
-			kind: "skill-invocation",
-			name: skill,
-			form: "instructions"
-		}
-	}));
-	return {
-		kind: "success",
-		text: `Queued ${skill} for the agent.`
-	};
-}
-function registerCommands(deps, commandCtx) {
-	commandCtx.commands.register({
-		name: "ponytail",
-		description: "Set or show Ponytail lazy senior dev intensity",
-		input: { hint: "[status|default <mode>|lite|full|ultra|off]" },
-		handler: ({ agent, rawInput }) => {
-			const input = rawInput.trim().toLowerCase();
-			const [head, ...rest] = input.split(/\s+/).filter(Boolean);
-			if ((head ?? "") === "default") {
-				let written;
-				try {
-					written = writeDefaultMode(rest[0]);
-				} catch (error) {
-					return {
-						kind: "error",
-						text: `Failed to save default: ${error.message}`
-					};
-				}
-				if (!written) return {
-					kind: "error",
-					text: "Usage: /ponytail default [lite|full|ultra|off]"
-				};
-				const effective = readDefaultMode(process.env, deps.profileMode);
-				deps.setDefault(effective);
-				const reason = defaultOverrideReason(process.env, deps.profileMode);
-				if (reason !== null) {
-					agent.steer(createUserMessage({
-						content: [{
-							type: "text",
-							text: `PONYTAIL DEFAULT SET — saved ${written}, effective ${effective} (${reason}).`
-						}],
-						source: {
-							kind: "plugin",
-							plugin: name
-						}
-					}));
-					return {
-						kind: "success",
-						text: `Saved default: ${written}. Effective default: ${effective}, overridden by ${reason}.`
-					};
-				}
-				agent.steer(createUserMessage({
-					content: [{
-						type: "text",
-						text: `PONYTAIL DEFAULT SET — new sessions start in ${written}.`
-					}],
-					source: {
-						kind: "plugin",
-						plugin: name
-					}
-				}));
+/**
+* The never-cut list. Every non-`off` mode keeps these; intensities tune how
+* aggressively code is minimized, never what may be dropped.
+*/
+const SAFETY_BOUNDARIES = [
+	"Never cut, in any mode:",
+	"- Input validation at trust boundaries.",
+	"- Error handling that prevents data loss.",
+	"- Security measures.",
+	"- Accessibility basics.",
+	"- Explicit acceptance criteria the user asked for.",
+	"- Understanding the problem and tracing the real flow first.",
+	"- The real end-to-end data flow: no UI-only field, unused state, placeholder path, or disconnected payload.",
+	"- Non-trivial logic leaves one minimal runnable check: the smallest assert or test that catches breakage; no framework or fixtures unless asked.",
+	"- Root-cause fixes over symptom patches.",
+	"- \"Minimal diff\" is not a substitute for \"correct fix\"."
+].join("\n");
+/** Lite: complete the explicit ask; reuse; suggest, do not challenge. */
+const LITE_RULES = [
+	"Execute the direct request without ceremony; complete every explicit acceptance criterion.",
+	"You may mention a simpler alternative briefly, but do not challenge or reject an explicit requirement.",
+	"Do not change the existing architecture merely to reduce line count.",
+	"Keep the smallest reasonable validation for non-trivial changes."
+].join("\n");
+/** Smallest complete end-to-end change: shared by Full and Ultra. */
+const E2E_RULES = [
+	"Smallest complete end-to-end change:",
+	"- Prefer the smallest complete end-to-end change compatible with the existing architecture, not merely the fewest lines in one file.",
+	"- Before creating a component, abstraction, protocol, migration, transport format, storage format, or dependency, inspect the repository’s existing path and preserve its current contract.",
+	"- Do not redesign transport, storage, API shape, or persistence when the task only asks for a local UI or behavior change.",
+	"- Prefer the smallest complete change across the real data flow: input → state → validation → payload → API → persistence → response/UI.",
+	"- It does not mean every layer must change: the change must be complete across the layers it touches.",
+	"- Do not leave a UI-only field, unused state, placeholder path, or disconnected payload merely because it produces a smaller diff."
+].join("\n");
+const MODE_RULES = {
+	lite: LITE_RULES,
+	full: [
+		"Use the seven-rung ladder; stop at the first rung that holds:",
+		"1. Does this need to exist at all? (YAGNI)",
+		"2. Does it already exist in this codebase? Reuse it.",
+		"3. Does the standard library do it? Use it.",
+		"4. Does a native platform feature cover it? Use it.",
+		"5. Does an already-installed dependency solve it? Use it.",
+		"6. Can the solution be reduced to a small expression? Make it that small.",
+		"7. Only then: write the minimum new implementation.",
+		"Default to the shortest correct implementation; prefer deletion and reuse. Fix root causes, not symptoms (one shared guard beats one per caller).",
+		E2E_RULES
+	].join("\n"),
+	ultra: [
+		"Require evidence before adding. Prefer deletion or reuse; challenge speculative features, caches, abstractions, configuration, migrations, and dependencies.",
+		"For complex requests, ship the smallest correct complete version and state what would justify a larger version.",
+		"Ultra is not refusal: explicit requirements and the safety boundaries remain mandatory.",
+		E2E_RULES
+	].join("\n")
+};
+const MODE_LABELS = {
+	lite: "Lite",
+	full: "Full",
+	ultra: "Ultra"
+};
+/** Compose the complete section text for one intensity. */
+function render(effective) {
+	return [
+		`PONYTAIL MODE ACTIVE — level: ${effective}`,
+		"",
+		INTRO,
+		"",
+		"## Common rules (all modes)",
+		COMMON_RULES,
+		"",
+		"## Safety boundaries (never cut)",
+		SAFETY_BOUNDARIES,
+		"",
+		`## ${MODE_LABELS[effective]} rules`,
+		MODE_RULES[effective]
+	].join("\n");
+}
+/**
+* The injected ruleset for one intensity, composed from the structured
+* fragments above. Returns an empty string for `off` (ponytail contributes
+* nothing). Renders are pure per mode and cached so every turn's bytes stay
+* identical.
+*/
+function getPonytailInstructions(mode) {
+	const effective = normalizeRuntimeMode(mode) ?? "full";
+	if (effective === "off") return "";
+	const cached = instructionCache.get(effective);
+	if (cached !== void 0) return cached;
+	const rendered = render(effective);
+	instructionCache.set(effective, rendered);
+	return rendered;
+}
+/** Rendered rulesets are pure per mode; cache to keep every turn's bytes identical. */
+const instructionCache = /* @__PURE__ */ new Map();
+//#endregion
+//#region lib/types/index.js
+/**
+* Ponytail: the "lazy senior developer" persona as a DeepSeek Harness plugin.
+*
+* One system-prompt section injects the mode-filtered ruleset every turn (the
+* always-on adapter), six runtime skills surface the review/audit/debt/gain/
+* help one-shots, six slash commands drive them from the command plane, and an
+* `agent/pre-step` listener honors the plain-text deactivation phrases.
+*
+* Mode is session-scoped and held in memory; the configured default resolves
+* from `PONYTAIL_DEFAULT_MODE`, then the Cordis profile `defaultMode`, then
+* `~/.config/ponytail/config.json` (see {@link readDefaultMode}), then
+* `full`. A session override via `/ponytail` outranks all of them.
+*
+* @module @mengyuly/dsh-ponytail
+*/
+const name = "ponytail";
+const inject = ["systemPrompt", "skills"];
+/** Prompt-section order: after the deployment persona (0), before tool guidance (100–199). */
+const SECTION_ORDER = 40;
+/** Build the one text-line notification a mode switch leaves for the model. */
+function modeNotice(mode) {
+	return mode === "off" ? "PONYTAIL MODE OFF" : `PONYTAIL MODE CHANGED — level: ${mode}`;
+}
+/** Extract the plain text of one user message (only its text blocks). */
+function messageText(message) {
+	const parts = [];
+	for (const block of message.content) if (block.type === "text" && typeof block.text === "string") parts.push(block.text);
+	return parts.join("\n");
+}
+/** Whether any message in a claimed batch is exactly a deactivation command. */
+function containsDeactivation(messages) {
+	return messages.some((message) => isDeactivationCommand(messageText(message)));
+}
+/** Mode visible to one agent: its session override, else the configured default. */
+function modeFor(deps, agent) {
+	return deps.store.modeFor(sessionKey(agent), deps.defaultMode());
+}
+/**
+* Queue one skill's full `<skill_content>` rendering as the model's next
+* ordinary turn, with the same user-explicit `skill-invocation` source the
+* built-in gesture boundary uses.
+*/
+async function queueSkill(deps, invocation, skill) {
+	const loaded = await deps.ctx.skills.get(skill, {
+		cwd: invocation.agent.session.header.cwd,
+		signal: invocation.signal
+	});
+	if (loaded === void 0) return {
+		kind: "error",
+		text: `skill "${skill}" is not available`
+	};
+	const notes = invocation.rawInput.trim();
+	const text = renderSkillContent(loaded) + (notes === "" ? "" : `\n\n${notes}`);
+	invocation.agent.followup(createUserMessage({
+		content: [{
+			type: "text",
+			text
+		}],
+		source: {
+			kind: "skill-invocation",
+			name: skill,
+			form: "instructions"
+		}
+	}));
+	return {
+		kind: "success",
+		text: `Queued ${skill} for the agent.`
+	};
+}
+function registerCommands(deps, commandCtx) {
+	commandCtx.commands.register({
+		name: "ponytail",
+		description: "Set or show Ponytail lazy senior dev intensity",
+		input: { hint: "[status|reset|default <mode>|lite|full|ultra|off]" },
+		handler: ({ agent, rawInput }) => {
+			const input = rawInput.trim().toLowerCase();
+			const [head, ...rest] = input.split(/\s+/).filter(Boolean);
+			if ((head ?? "") === "default") {
+				let written;
+				try {
+					written = writeDefaultMode(rest[0]);
+				} catch (error) {
+					return {
+						kind: "error",
+						text: `Failed to save default: ${error.message}`
+					};
+				}
+				if (!written) return {
+					kind: "error",
+					text: "Usage: /ponytail default [lite|full|ultra|off]"
+				};
+				const effective = readDefaultMode(process.env, deps.profileMode);
+				deps.setDefault(effective);
+				const reason = defaultOverrideReason(process.env, deps.profileMode);
+				if (reason !== null) {
+					agent.steer(createUserMessage({
+						content: [{
+							type: "text",
+							text: `PONYTAIL DEFAULT SET — saved ${written}, effective ${effective} (${reason}).`
+						}],
+						source: {
+							kind: "plugin",
+							plugin: name
+						}
+					}));
+					return {
+						kind: "success",
+						text: `Saved default: ${written}. Effective default: ${effective}, overridden by ${reason}.`
+					};
+				}
+				agent.steer(createUserMessage({
+					content: [{
+						type: "text",
+						text: `PONYTAIL DEFAULT SET — new sessions start in ${written}.`
+					}],
+					source: {
+						kind: "plugin",
+						plugin: name
+					}
+				}));
+				return {
+					kind: "success",
+					text: `Ponytail default set — new sessions start in ${written}.`
+				};
+			}
+			if (input === "status") {
+				const current = modeFor(deps, agent);
+				const source = deps.store.has(sessionKey(agent)) ? "session override" : "configured default";
 				return {
 					kind: "success",
-					text: `Ponytail default set — new sessions start in ${written}.`
+					text: `Ponytail mode: ${current} (${source}). Use /ponytail reset|lite|full|ultra|off.`
 				};
 			}
-			if (input === "status") return {
-				kind: "success",
-				text: `Ponytail mode: ${modeFor(deps, agent)}. Use /ponytail lite|full|ultra|off.`
-			};
-			if (input === "") {
-				const current = modeFor(deps, agent);
-				const effectiveDefault = deps.defaultMode();
-				if (current === "off") {
-					if (effectiveDefault === "off") {
-						deps.store.set(sessionKey(agent), "full");
-						agent.steer(createUserMessage({
-							content: [{
-								type: "text",
-								text: "PONYTAIL MODE CHANGED — level: full"
-							}],
-							source: {
-								kind: "plugin",
-								plugin: name
-							}
-						}));
-						return {
-							kind: "success",
-							text: "Ponytail re-enabled at full (the effective default is off)."
-						};
-					}
-					deps.store.clear(sessionKey(agent));
-					agent.steer(createUserMessage({
-						content: [{
-							type: "text",
-							text: `PONYTAIL MODE ACTIVE — level: ${effectiveDefault}`
-						}],
-						source: {
-							kind: "plugin",
-							plugin: name
-						}
-					}));
-					return {
-						kind: "success",
-						text: `Ponytail re-enabled. Effective default: ${effectiveDefault}.`
-					};
-				}
-				agent.steer(createUserMessage({
+			if (input === "reset") {
+				const key = sessionKey(agent);
+				const changed = deps.store.has(key);
+				deps.store.clear(key);
+				const current = deps.defaultMode();
+				if (changed) agent.steer(createUserMessage({
 					content: [{
 						type: "text",
-						text: `PONYTAIL MODE ACTIVE — level: ${current}`
+						text: modeNotice(current)
 					}],
 					source: {
 						kind: "plugin",
@@ -1798,139 +1751,191 @@ function registerCommands(deps, commandCtx) {
 				}));
 				return {
 					kind: "success",
-					text: `Ponytail mode: ${current}. Use /ponytail lite|full|ultra|off.`
+					text: changed ? `Ponytail session override cleared. Effective mode: ${current}.` : `Ponytail already follows the configured default: ${current}.`
 				};
 			}
-			const mode = normalizeRuntimeMode(input);
-			if (!mode) return {
-				kind: "error",
-				text: "Usage: /ponytail [status|default <mode>|lite|full|ultra|off]"
-			};
-			deps.store.set(sessionKey(agent), mode);
-			agent.steer(createUserMessage({
-				content: [{
-					type: "text",
-					text: modeNotice(mode)
-				}],
-				source: {
-					kind: "plugin",
-					plugin: name
-				}
-			}));
-			return {
-				kind: "success",
-				text: mode === "off" ? "Ponytail mode off." : `Ponytail mode set to ${mode}.`
-			};
-		}
-	});
-	for (const skill of [
-		"ponytail-review",
-		"ponytail-audit",
-		"ponytail-debt",
-		"ponytail-gain",
-		"ponytail-help"
-	]) commandCtx.commands.register({
-		name: skill,
-		description: descriptionFor(skill),
-		input: { hint: "[notes]" },
-		handler: (invocation) => queueSkill(deps, invocation, skill)
-	});
-}
-/** One-line command catalog copy, kept beside the skills for discovery parity. */
-function descriptionFor(skill) {
-	switch (skill) {
-		case "ponytail-review": return "Over-engineering review of the current changes";
-		case "ponytail-audit": return "Whole-repo over-engineering audit (what can be deleted)";
-		case "ponytail-debt": return "Harvest ponytail: comments into a tracked debt ledger";
-		case "ponytail-gain": return "Show ponytail measured-impact scoreboard (less code, cost, time)";
-		case "ponytail-help": return "Quick reference for ponytail levels, skills, and commands";
-		default: return `Run the ${skill} skill`;
-	}
-}
-/**
-* Register the always-on ruleset section, the runtime skills, the slash
-* commands, and the plain-text deactivation listener.
-*/
-function apply(ctx, config = {}) {
-	const profileMode = normalizeRuntimeMode(config.defaultMode);
-	if (config.defaultMode !== void 0 && profileMode === null) ctx.logger.warn(`[ponytail] profile config defaultMode is not lite|full|ultra|off: ${JSON.stringify(config.defaultMode)}; falling back`);
-	let defaultMode = null;
-	const warned = /* @__PURE__ */ new Set();
-	const warnOnce = (key, message) => {
-		if (warned.has(key)) return;
-		warned.add(key);
-		ctx.logger.warn(`[ponytail] ${message}`);
-	};
-	const refreshDefault = () => {
-		const resolution = readDefaultModeInfo(process.env, profileMode);
-		if (resolution.issue) warnOnce(`default:${resolution.issue.kind}`, `${resolution.issue.detail}; using ${resolution.mode}`);
-		defaultMode = resolution.mode;
-		return defaultMode;
-	};
-	const readDefault = () => defaultMode ?? refreshDefault();
-	const setDefault = (mode) => {
-		defaultMode = mode;
-	};
-	const store = new ModeStore();
-	const matcherResult = compileSubagentMatcher(process.env.PONYTAIL_SUBAGENT_MATCHER);
-	const matcher = matcherResult.matcher;
-	if (matcherResult.invalid) warnOnce("matcher:invalid", "PONYTAIL_SUBAGENT_MATCHER is not a valid regular expression; ignoring it (fail-open).");
-	const configFile = configPath();
-	const onConfigChange = () => {
-		const resolution = readDefaultModeInfo(process.env, profileMode);
-		if (resolution.issue) {
-			warnOnce(`config:${resolution.issue.kind}`, `${resolution.issue.detail}; keeping the previous default`);
-			return;
-		}
-		defaultMode = resolution.mode;
-	};
-	watchFile(configFile, { interval: 1e3 }, onConfigChange).unref();
-	ctx.effect(() => () => {
-		unwatchFile(configFile, onConfigChange);
-	}, "ponytail: config hot reload");
-	ctx.on("agent/disposed", ({ agent }) => {
-		store.clear(sessionKey(agent));
-	});
-	ctx.systemPrompt.section({
-		name: "ponytail",
-		order: SECTION_ORDER,
-		text: ({ agent }) => {
-			if (agent && matcher && isSubagentSession(agent.session.header)) {
-				const preset = agent.session.header.agentPreset;
-				if (preset && !matcher.test(preset)) return "";
-			}
-			return getPonytailInstructions(agent ? store.modeFor(sessionKey(agent), readDefault()) : readDefault());
-		}
-	});
-	for (const skill of ponytailSkills()) ctx.skills.register(skill);
-	ctx.inject(["commands"], (commandCtx) => {
-		registerCommands({
-			ctx,
-			store,
-			profileMode,
-			defaultMode: readDefault,
-			setDefault
-		}, commandCtx);
-	});
-	ctx.on("agent/pre-step", async (payload, next) => {
-		const deactivated = containsDeactivation(payload.messages);
-		if (deactivated) store.set(sessionKey(payload.agent), "off");
-		const decision = await next();
-		if (deactivated && decision.kind === "enter") return {
-			kind: "enter",
-			messages: [...decision.messages, createUserMessage({
-				content: [{
-					type: "text",
-					text: "PONYTAIL MODE OFF"
-				}],
-				source: {
-					kind: "plugin",
-					plugin: name
-				}
-			})]
-		};
-		return decision;
-	});
-}
-//#endregion
-export { apply, containsDeactivation, inject, messageText, name };
+			if (input === "") {
+				const current = modeFor(deps, agent);
+				const effectiveDefault = deps.defaultMode();
+				if (current === "off") {
+					if (effectiveDefault === "off") {
+						deps.store.set(sessionKey(agent), "full");
+						agent.steer(createUserMessage({
+							content: [{
+								type: "text",
+								text: "PONYTAIL MODE CHANGED — level: full"
+							}],
+							source: {
+								kind: "plugin",
+								plugin: name
+							}
+						}));
+						return {
+							kind: "success",
+							text: "Ponytail re-enabled at full (the effective default is off)."
+						};
+					}
+					deps.store.clear(sessionKey(agent));
+					agent.steer(createUserMessage({
+						content: [{
+							type: "text",
+							text: `PONYTAIL MODE ACTIVE — level: ${effectiveDefault}`
+						}],
+						source: {
+							kind: "plugin",
+							plugin: name
+						}
+					}));
+					return {
+						kind: "success",
+						text: `Ponytail re-enabled. Effective default: ${effectiveDefault}.`
+					};
+				}
+				agent.steer(createUserMessage({
+					content: [{
+						type: "text",
+						text: `PONYTAIL MODE ACTIVE — level: ${current}`
+					}],
+					source: {
+						kind: "plugin",
+						plugin: name
+					}
+				}));
+				return {
+					kind: "success",
+					text: `Ponytail mode: ${current}. Use /ponytail reset|lite|full|ultra|off.`
+				};
+			}
+			const mode = normalizeRuntimeMode(input);
+			if (!mode) return {
+				kind: "error",
+				text: "Usage: /ponytail [status|reset|default <mode>|lite|full|ultra|off]"
+			};
+			deps.store.set(sessionKey(agent), mode);
+			agent.steer(createUserMessage({
+				content: [{
+					type: "text",
+					text: modeNotice(mode)
+				}],
+				source: {
+					kind: "plugin",
+					plugin: name
+				}
+			}));
+			return {
+				kind: "success",
+				text: mode === "off" ? "Ponytail mode off." : `Ponytail mode set to ${mode}.`
+			};
+		}
+	});
+	for (const skill of [
+		"ponytail-review",
+		"ponytail-audit",
+		"ponytail-debt",
+		"ponytail-gain",
+		"ponytail-help"
+	]) commandCtx.commands.register({
+		name: skill,
+		description: descriptionFor(skill),
+		input: { hint: "[notes]" },
+		handler: (invocation) => queueSkill(deps, invocation, skill)
+	});
+}
+/** One-line command catalog copy, kept beside the skills for discovery parity. */
+function descriptionFor(skill) {
+	switch (skill) {
+		case "ponytail-review": return "Over-engineering review of the current changes";
+		case "ponytail-audit": return "Whole-repo over-engineering audit (what can be deleted)";
+		case "ponytail-debt": return "Harvest ponytail: comments into a tracked debt ledger";
+		case "ponytail-gain": return "Show ponytail measured-impact scoreboard (less code, cost, time)";
+		case "ponytail-help": return "Quick reference for ponytail levels, skills, and commands";
+		default: return `Run the ${skill} skill`;
+	}
+}
+/**
+* Register the always-on ruleset section, the runtime skills, the slash
+* commands, and the plain-text deactivation listener.
+*/
+function apply(ctx, config = {}) {
+	const profileMode = normalizeRuntimeMode(config.defaultMode);
+	if (config.defaultMode !== void 0 && profileMode === null) ctx.logger.warn(`[ponytail] profile config defaultMode is not lite|full|ultra|off: ${JSON.stringify(config.defaultMode)}; falling back`);
+	let defaultMode = null;
+	const warned = /* @__PURE__ */ new Set();
+	const warnOnce = (key, message) => {
+		if (warned.has(key)) return;
+		warned.add(key);
+		ctx.logger.warn(`[ponytail] ${message}`);
+	};
+	const refreshDefault = () => {
+		const resolution = readDefaultModeInfo(process.env, profileMode);
+		if (resolution.issue) warnOnce(`default:${resolution.issue.kind}`, `${resolution.issue.detail}; using ${resolution.mode}`);
+		defaultMode = resolution.mode;
+		return defaultMode;
+	};
+	const readDefault = () => defaultMode ?? refreshDefault();
+	const setDefault = (mode) => {
+		defaultMode = mode;
+	};
+	const store = new ModeStore();
+	const matcherResult = compileSubagentMatcher(process.env.PONYTAIL_SUBAGENT_MATCHER);
+	const matcher = matcherResult.matcher;
+	if (matcherResult.invalid) warnOnce("matcher:invalid", "PONYTAIL_SUBAGENT_MATCHER is not a valid regular expression; ignoring it (fail-open).");
+	const configFile = configPath();
+	const onConfigChange = () => {
+		const resolution = readDefaultModeInfo(process.env, profileMode);
+		if (resolution.issue) {
+			warnOnce(`config:${resolution.issue.kind}`, `${resolution.issue.detail}; keeping the previous default`);
+			return;
+		}
+		defaultMode = resolution.mode;
+	};
+	watchFile(configFile, { interval: 1e3 }, onConfigChange).unref();
+	ctx.effect(() => () => {
+		unwatchFile(configFile, onConfigChange);
+	}, "ponytail: config hot reload");
+	ctx.on("agent/disposed", ({ agent }) => {
+		store.clear(sessionKey(agent));
+	});
+	ctx.systemPrompt.section({
+		name: "ponytail",
+		order: SECTION_ORDER,
+		text: ({ agent }) => {
+			if (agent && matcher && isSubagentSession(agent.session.header)) {
+				const preset = agent.session.header.agentPreset;
+				if (preset && !matcher.test(preset)) return "";
+			}
+			return getPonytailInstructions(agent ? store.modeFor(sessionKey(agent), readDefault()) : readDefault());
+		}
+	});
+	for (const skill of ponytailSkills()) ctx.skills.register(skill);
+	ctx.inject(["commands"], (commandCtx) => {
+		registerCommands({
+			ctx,
+			store,
+			profileMode,
+			defaultMode: readDefault,
+			setDefault
+		}, commandCtx);
+	});
+	ctx.on("agent/pre-step", async (payload, next) => {
+		const deactivated = containsDeactivation(payload.messages);
+		if (deactivated) store.set(sessionKey(payload.agent), "off");
+		const decision = await next();
+		if (deactivated && decision.kind === "enter") return {
+			kind: "enter",
+			messages: [...decision.messages, createUserMessage({
+				content: [{
+					type: "text",
+					text: "PONYTAIL MODE OFF"
+				}],
+				source: {
+					kind: "plugin",
+					plugin: name
+				}
+			})]
+		};
+		return decision;
+	});
+}
+//#endregion
+export { apply, containsDeactivation, inject, messageText, name };