@graphorin/provider-llamacpp-node 0.6.0 → 0.7.0

package/CHANGELOG.md

@@ -1,5 +1,35 @@
 # @graphorin/provider-llamacpp-node
 
+## 0.7.0
+
+### Minor Changes
+
+- [#154](https://github.com/o-stepper/graphorin/pull/154) [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04) Thanks [@o-stepper](https://github.com/o-stepper)! - W-072: every export map's `import` condition becomes `default`, and the Node floor rises to `>=22.12.0`.
+
+  CJS consumers previously hit a bewildering `ERR_PACKAGE_PATH_NOT_EXPORTED` instead of a clear ESM-only signal. With the `default` condition, plain `require('@graphorin/core')` works via Node's stable `require(esm)` - which shipped in 22.12, hence the engines bump across every workspace manifest (packages, examples, benchmarks, docs; enforced by the widened mvp-readiness sweep). No dual-instance hazard: there is no CJS build, `require()` returns the same ESM module instance. ESM consumers are unaffected (`default` serves both paths; `types` stays first). The pack gate now runs attw under the full `node16` profile (was `esm-only`) and adds a runtime `require(esm)` smoke against the packed tarballs. Installs on Node 22.0-22.11 with `engine-strict` will refuse - upgrade Node (see the migration guide).
+
+- [#160](https://github.com/o-stepper/graphorin/pull/160) [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156) Thanks [@o-stepper](https://github.com/o-stepper)! - W-096: `llamaCppNodeAdapter` speaks real chat history. When the session exposes node-llama-cpp v3's `setChatHistory` (the default factory's sessions do), a multi-turn transcript is fed as structured history turns (system included - `setChatHistory` replaces the whole history, so relying on the constructor `systemPrompt` alone would drop it) and only the LAST user turn is prompted - the pre-fix pseudo-prompt serialised the entire conversation into one `[user]`-framed string, double-templating it and degrading chat-tuned GGUF models. Sessions without the setter (older fixtures, custom factories) keep the legacy renderPrompt path byte-identically. `completionTokens` now tokenize the ASSEMBLED response once instead of per streamed chunk (BPE merges span chunk seams, so per-chunk counting over-billed). New opt-in `persistentSession: true` caches one session (context + KV cache) per adapter instance, serialised by a promise mutex with per-request `setChatHistory` re-sync, so an agent loop stops re-prefilling the growing transcript every step; the default per-request create/dispose lifecycle is unchanged.
+
+### Patch Changes
+
+- [#164](https://github.com/o-stepper/graphorin/pull/164) [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00) Thanks [@o-stepper](https://github.com/o-stepper)! - Tarballs now ship `src/` so the published `dist/**/*.d.ts.map` files actually work (W-136): the maps referenced `../src/*.ts` that the `files` whitelist excluded, so go-to-definition fell back into `.d.ts` and the shipped maps were dead weight. The pack gate gains a `map-integrity` leg: every source referenced by a shipped map must resolve inside the tarball (or be embedded via `sourcesContent`), with an anti-vacuous guard - a package whose tsdown config emits declaration maps must contain a non-zero number of `.d.ts.map` files, so a cache-restored dist that silently dropped maps fails the gate instead of passing vacuously. `mvp-readiness` now requires `src` in every publishable `files` array.
+
+- [#164](https://github.com/o-stepper/graphorin/pull/164) [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00) Thanks [@o-stepper](https://github.com/o-stepper)! - Every published package now declares its tree-shaking contract via `sideEffects` (W-137): 18 packages audited to a pure module scope get `false`, the CLI declares its bin entry (`["./dist/bin/*"]`), and `@graphorin/security` gets an explicit `true` - its secrets subsystem registers built-in resolvers and the SecretValue caller-context provider at import time, so marking it pure would let bundlers drop those registrations. `mvp-readiness` now fails any publishable manifest without a declared `sideEffects`, closing the drift for future packages.
+
+- Updated dependencies [[`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`832f22e`](https://github.com/o-stepper/graphorin/commit/832f22e570b8c3175c1adeb4c150070cbd131534), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`32bbd03`](https://github.com/o-stepper/graphorin/commit/32bbd03b588136a355e4b5ad6ac5e19b36b4d8ab), [`32bbd03`](https://github.com/o-stepper/graphorin/commit/32bbd03b588136a355e4b5ad6ac5e19b36b4d8ab), [`32bbd03`](https://github.com/o-stepper/graphorin/commit/32bbd03b588136a355e4b5ad6ac5e19b36b4d8ab), [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156), [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156), [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156), [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156), [`4ee256e`](https://github.com/o-stepper/graphorin/commit/4ee256e30fe9190cef6c48dc6785464757707156), [`32bbd03`](https://github.com/o-stepper/graphorin/commit/32bbd03b588136a355e4b5ad6ac5e19b36b4d8ab), [`32bbd03`](https://github.com/o-stepper/graphorin/commit/32bbd03b588136a355e4b5ad6ac5e19b36b4d8ab), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`764239b`](https://github.com/o-stepper/graphorin/commit/764239b97e0e0202442e91272583f13adeb12d00), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04), [`fe98522`](https://github.com/o-stepper/graphorin/commit/fe98522ce2477c9a7dc09029f9dcfdb1f7c9aa04)]:
+  - @graphorin/core@0.7.0
+  - @graphorin/provider@0.7.0
+
+## 0.6.1
+
+### Patch Changes
+
+- [#142](https://github.com/o-stepper/graphorin/pull/142) [`436d6ca`](https://github.com/o-stepper/graphorin/commit/436d6ca5ebbd16df094e915682d3915c279a8430) Thanks [@o-stepper](https://github.com/o-stepper)! - Version constants and version-bearing strings now derive from each package's manifest at build time (`VERSION = pkg.version`; writer ids, client/server info, OTLP framework attributes, build-info metrics interpolate it). No behavioral change at the current version: the rendered strings are byte-identical. A release bump no longer edits source; the new `check-version-consistency` gate fails any reintroduced hardcoded framework version.
+
+- Updated dependencies [[`436d6ca`](https://github.com/o-stepper/graphorin/commit/436d6ca5ebbd16df094e915682d3915c279a8430)]:
+  - @graphorin/core@0.6.1
+  - @graphorin/provider@0.6.1
+
 ## 0.6.0
 
 ### Patch Changes

package/README.md

@@ -78,5 +78,5 @@ discipline rather than enforcing it at runtime.
 
 ## Project metadata
 
-- **Project Graphorin** · v0.6.0 · MIT License · © 2026 Oleksiy Stepurenko
+- **Project Graphorin** · v0.7.0 · MIT License · © 2026 Oleksiy Stepurenko
 - Repository: <https://github.com/o-stepper/graphorin>

package/dist/adapter.d.ts

@@ -41,6 +41,19 @@ interface LlamaCppNodeAdapterOptions {
    * seam.
    */
   readonly sessionFactory?: (model: LlamaModelInstance, system?: string) => Promise<LlamaSessionInstance>;
+  /**
+   * W-096: reuse ONE session (context + KV cache) across requests
+   * instead of creating and disposing a fresh one per call - an agent
+   * loop then avoids re-prefilling the growing transcript on every
+   * step. Requests serialise through a promise mutex (a llama context
+   * sequence is single-threaded), and the chat history re-syncs via
+   * `setChatHistory` before each prompt. Strictly opt-in: the default
+   * per-request lifecycle stays memory-safe and concurrency-safe; the
+   * cached session also skips per-request disposal (it lives until the
+   * process / instance is released). Sessions WITHOUT `setChatHistory`
+   * cannot re-sync and silently degrade to per-request behaviour.
+   */
+  readonly persistentSession?: boolean;
 }
 /**
  * Build a Graphorin {@link Provider} backed by an in-process GGUF

package/dist/adapter.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"adapter.d.ts","names":[],"sources":["../src/adapter.ts"],"sourcesContent":[],"mappings":";;;;;;;;;;UA8CiB,0BAAA;;;;;;;;;;0BAUS,QAAQ;;gCAEF,cAAc;;;;;8BAKhB;;;;;;2BAMH;;;;;;;;;oCAUhB,wCAEJ,QAAQ;;;;;;;;;iBAsBC,mBAAA,UAA6B,6BAA6B"}
+{"version":3,"file":"adapter.d.ts","names":[],"sources":["../src/adapter.ts"],"sourcesContent":[],"mappings":";;;;;;;;;;UA+CiB,0BAAA;;;;;;;;;;0BAUS,QAAQ;;gCAEF,cAAc;;;;;8BAKhB;;;;;;2BAMH;;;;;;;;;oCAUhB,wCAEJ,QAAQ;;;;;;;;;;;;;;;;;;;;;;iBAmCC,mBAAA,UAA6B,6BAA6B"}

package/dist/adapter.js

@@ -51,17 +51,17 @@ function llamaCppNodeAdapter(options) {
 		})();
 		return resolving;
 	};
-	const sessionFactory = options.sessionFactory ?? options.runtimeOverrides?.createSession ?? ((model, system) => defaultSessionFactory(model, system, options));
+	const acquireSession = buildSessionManager(options, options.sessionFactory ?? options.runtimeOverrides?.createSession ?? ((model, system) => defaultSessionFactory(model, system, options)));
 	return {
 		name: providerName,
 		modelId: options.modelPath,
 		capabilities,
 		acceptsSensitivity,
 		stream(req) {
-			return streamLlamaCppNode(ensureModel, sessionFactory, providerName, options, applyLlamaCppNodePreflight(req, capabilities));
+			return streamLlamaCppNode(ensureModel, acquireSession, providerName, options, applyLlamaCppNodePreflight(req, capabilities));
 		},
 		async generate(req) {
-			const events = streamLlamaCppNode(ensureModel, sessionFactory, providerName, options, applyLlamaCppNodePreflight(req, capabilities));
+			const events = streamLlamaCppNode(ensureModel, acquireSession, providerName, options, applyLlamaCppNodePreflight(req, capabilities));
 			const collected = [];
 			let usage = {
 				promptTokens: 0,
@@ -107,11 +107,18 @@ function applyLlamaCppNodePreflight(req, capabilities) {
 		messages: filtered
 	};
 }
-async function* streamLlamaCppNode(ensureModel, sessionFactory, providerName, options, req) {
+async function* streamLlamaCppNode(ensureModel, acquireSession, providerName, options, req) {
 	const { model } = await ensureModel();
-	const session = await sessionFactory(model, typeof req.systemMessage === "string" ? req.systemMessage : void 0);
-	const prompt = renderPrompt(req);
-	const promptTokens = lengthOf(model.tokenize(prompt));
+	const lease = await acquireSession(model, typeof req.systemMessage === "string" ? req.systemMessage : void 0);
+	const session = lease.session;
+	const historic = typeof session.setChatHistory === "function";
+	let prompt;
+	if (historic) {
+		const { priorTurns, lastUserText } = buildChatHistory(req);
+		session.setChatHistory?.(priorTurns);
+		prompt = lastUserText;
+	} else prompt = renderPrompt(req);
+	const promptTokens = lengthOf(model.tokenize(renderPrompt(req)));
 	yield {
 		type: "stream-start",
 		metadata: {
@@ -120,7 +127,7 @@ async function* streamLlamaCppNode(ensureModel, sessionFactory, providerName, op
 			createdAt: (/* @__PURE__ */ new Date()).toISOString()
 		}
 	};
-	let completionTokens = 0;
+	const pieces = [];
 	let errored = false;
 	let aborted = false;
 	try {
@@ -134,7 +141,7 @@ async function* streamLlamaCppNode(ensureModel, sessionFactory, providerName, op
 				break;
 			}
 			if (typeof piece !== "string" || piece.length === 0) continue;
-			completionTokens += lengthOf(model.tokenize(piece));
+			pieces.push(piece);
 			yield {
 				type: "text-delta",
 				delta: piece
@@ -154,10 +161,9 @@ async function* streamLlamaCppNode(ensureModel, sessionFactory, providerName, op
 			}
 		};
 	} finally {
-		try {
-			session.dispose?.();
-		} catch {}
+		lease.release();
 	}
+	const completionTokens = lengthOf(model.tokenize(pieces.join("")));
 	yield {
 		type: "finish",
 		finishReason: errored ? "error" : aborted ? "aborted" : "stop",
@@ -168,6 +174,103 @@ async function* streamLlamaCppNode(ensureModel, sessionFactory, providerName, op
 		}
 	};
 }
+/**
+* W-096: per-request sessions by default (create -> use -> dispose, the
+* memory-safe core-provider-08 lifecycle); with `persistentSession:
+* true` ONE cached session serialised by a promise mutex (a llama
+* context sequence is single-threaded) whose history re-syncs via
+* `setChatHistory` on every request. A cached session that turns out
+* not to support `setChatHistory` degrades to per-request behaviour -
+* without the setter its history cannot be made consistent.
+*/
+function buildSessionManager(options, sessionFactory) {
+	if (options.persistentSession !== true) return async (model, system) => {
+		const session = await sessionFactory(model, system);
+		return {
+			session,
+			release: () => {
+				try {
+					session.dispose?.();
+				} catch {}
+			}
+		};
+	};
+	let cached = null;
+	let queueTail = Promise.resolve();
+	return async (model, system) => {
+		let unlock;
+		const gate = new Promise((resolve) => {
+			unlock = resolve;
+		});
+		const priorTail = queueTail;
+		queueTail = queueTail.then(() => gate);
+		await priorTail;
+		if (cached === null) cached = await sessionFactory(model, system);
+		const session = cached;
+		if (typeof session.setChatHistory !== "function") {
+			cached = null;
+			return {
+				session,
+				release: () => {
+					try {
+						session.dispose?.();
+					} catch {}
+					unlock();
+				}
+			};
+		}
+		return {
+			session,
+			release: () => unlock()
+		};
+	};
+}
+/**
+* W-096: split the request into node-llama-cpp chat-history turns plus
+* the trailing user text (the prompt). The system message is included
+* as a history item because `setChatHistory` REPLACES the session's
+* whole history - including the slot the constructor `systemPrompt`
+* seeded - so relying on the constructor alone would drop it. Tool
+* roles cannot occur (`toolCalling: false`); a defensive skip keeps a
+* stray one from corrupting the template.
+*/
+function buildChatHistory(req) {
+	const turns = [];
+	if (typeof req.systemMessage === "string" && req.systemMessage.length > 0) turns.push({
+		type: "system",
+		text: req.systemMessage
+	});
+	for (const msg of req.messages) {
+		const text = textOf(msg);
+		if (msg.role === "user") turns.push({
+			type: "user",
+			text
+		});
+		else if (msg.role === "assistant") turns.push({
+			type: "model",
+			response: [text]
+		});
+		else if (msg.role === "system") turns.push({
+			type: "system",
+			text
+		});
+	}
+	const last = turns[turns.length - 1];
+	if (last !== void 0 && last.type === "user") {
+		turns.pop();
+		return {
+			priorTurns: turns,
+			lastUserText: last.text
+		};
+	}
+	return {
+		priorTurns: turns,
+		lastUserText: ""
+	};
+}
+function textOf(msg) {
+	return typeof msg.content === "string" ? msg.content : msg.content.map((p) => p.type === "text" ? p.text : p.type === "reasoning" ? p.text : "").join("");
+}
 async function resolveModel(options) {
 	return (await loadLlamaModule(options.runtimeOverrides)).loadModel({
 		modelPath: options.modelPath,

package/dist/adapter.js.map

@@ -1 +1 @@
-{"version":3,"file":"adapter.js","names":["LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY: ReadonlyArray<Sensitivity>","DEFAULT_CAPABILITIES: ProviderCapabilities","capabilities: ProviderCapabilities","resolved: { model: LlamaModelInstance } | null","resolving: Promise<{ model: LlamaModelInstance }> | null","collected: string[]","usage: Usage","finishReason: ProviderResponse['finishReason']","streamError: { message: string } | undefined","streamOptions: { signal?: AbortSignal; maxTokens?: number; temperature?: number }","queue: string[]","failure: unknown","wake: (() => void) | null","parts: string[]"],"sources":["../src/adapter.ts"],"sourcesContent":["/**\n * `llamaCppNodeAdapter` - in-process GGUF adapter built on\n * `node-llama-cpp@^3.5`. The adapter declares `trust: 'loopback'`\n * permanently because the model lives in the same trust boundary as\n * the host process.\n *\n * @packageDocumentation\n */\n\nimport type {\n  Provider,\n  ProviderCapabilities,\n  ProviderEvent,\n  ProviderRequest,\n  ProviderResponse,\n  Sensitivity,\n  Usage,\n} from '@graphorin/core';\nimport { ProviderHttpError } from '@graphorin/provider/errors';\nimport { applyReasoningPolicy, resolveReasoningRetention } from '@graphorin/provider/reasoning';\nimport {\n  type LlamaChatSessionPeer,\n  type LlamaCppNodeRuntimeOverrides,\n  type LlamaModelInstance,\n  type LlamaSessionInstance,\n  loadLlamaChatSessionCtor,\n  loadLlamaModule,\n} from './runtime.js';\n\n/**\n * Default sensitivity envelope for the in-process adapter - same as\n * the loopback HTTP path.\n *\n * @stable\n */\nexport const LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY: ReadonlyArray<Sensitivity> = Object.freeze([\n  'public',\n  'internal',\n  'secret',\n]);\n\n/**\n * Options accepted by {@link llamaCppNodeAdapter}.\n *\n * @stable\n */\nexport interface LlamaCppNodeAdapterOptions {\n  /** Filesystem path to the `.gguf` model file. */\n  readonly modelPath: string;\n  /** Number of layers to offload to the GPU. Default `'auto'`. */\n  readonly gpuLayers?: number | 'auto';\n  /** Optional context-window override. */\n  readonly contextSize?: number;\n  /** Provider name attached to spans / log lines. */\n  readonly name?: string;\n  /** Capability declaration. Merged on top of the defaults table. */\n  readonly capabilities?: Partial<ProviderCapabilities>;\n  /** Sensitivity override (defaults to the loopback envelope). */\n  readonly acceptsSensitivity?: ReadonlyArray<Sensitivity>;\n  /**\n   * Test-only runtime override. When unset the adapter loads\n   * `node-llama-cpp` lazily on first call.\n   */\n  readonly runtimeOverrides?: LlamaCppNodeRuntimeOverrides;\n  /**\n   * Optional `model` override that short-circuits the\n   * `loadLlamaModule(...).loadModel(...)` flow. Tests pass a fixture\n   * shaped instance.\n   */\n  readonly modelOverride?: LlamaModelInstance;\n  /**\n   * Optional session factory override. When unset, the adapter builds a\n   * real session from the peer (PS-3): `model.createContext()` →\n   * `new LlamaChatSession({ contextSequence })`, streaming through\n   * `prompt(text, { onTextChunk })`. Overrides\n   * (`runtimeOverrides.createSession` or this option) keep the test\n   * seam.\n   */\n  readonly sessionFactory?: (\n    model: LlamaModelInstance,\n    system?: string,\n  ) => Promise<LlamaSessionInstance>;\n}\n\nconst DEFAULT_CAPABILITIES: ProviderCapabilities = {\n  streaming: true,\n  toolCalling: false,\n  parallelToolCalls: false,\n  multimodal: false,\n  structuredOutput: false,\n  reasoning: false,\n  contextWindow: 8_192,\n  maxOutput: 4_096,\n  reasoningContract: 'optional',\n};\n\n/**\n * Build a Graphorin {@link Provider} backed by an in-process GGUF\n * model. The first call lazily loads the `node-llama-cpp` peer + the\n * model file; subsequent calls reuse the cached instances.\n *\n * @stable\n */\nexport function llamaCppNodeAdapter(options: LlamaCppNodeAdapterOptions): Provider {\n  const providerName = options.name ?? `llama-cpp-node-${basename(options.modelPath)}`;\n  const capabilities: ProviderCapabilities = {\n    ...DEFAULT_CAPABILITIES,\n    ...options.capabilities,\n  };\n  const acceptsSensitivity = options.acceptsSensitivity ?? LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY;\n  let resolved: { model: LlamaModelInstance } | null = null;\n  let resolving: Promise<{ model: LlamaModelInstance }> | null = null;\n  const ensureModel = async (): Promise<{ model: LlamaModelInstance }> => {\n    if (resolved !== null) return resolved;\n    if (resolving !== null) return resolving;\n    resolving = (async () => {\n      const model = options.modelOverride ?? (await resolveModel(options));\n      resolved = { model };\n      resolving = null;\n      return resolved;\n    })();\n    return resolving;\n  };\n  const sessionFactory =\n    options.sessionFactory ??\n    options.runtimeOverrides?.createSession ??\n    ((model: LlamaModelInstance, system?: string) => defaultSessionFactory(model, system, options));\n  return {\n    name: providerName,\n    modelId: options.modelPath,\n    capabilities,\n    acceptsSensitivity,\n    stream(req) {\n      return streamLlamaCppNode(\n        ensureModel,\n        sessionFactory,\n        providerName,\n        options,\n        applyLlamaCppNodePreflight(req, capabilities),\n      );\n    },\n    async generate(req) {\n      const events = streamLlamaCppNode(\n        ensureModel,\n        sessionFactory,\n        providerName,\n        options,\n        applyLlamaCppNodePreflight(req, capabilities),\n      );\n      const collected: string[] = [];\n      let usage: Usage = { promptTokens: 0, completionTokens: 0, totalTokens: 0 };\n      let finishReason: ProviderResponse['finishReason'] = 'stop';\n      let streamError: { message: string } | undefined;\n      for await (const event of events) {\n        if (event.type === 'text-delta') collected.push(event.delta);\n        if (event.type === 'error') streamError = event.error;\n        if (event.type === 'finish') {\n          usage = event.usage;\n          finishReason = event.finishReason;\n        }\n      }\n      // PS-4: a swallowed mid-stream error returned truncated text\n      // indistinguishable from success - and a never-throwing\n      // generate() bypasses withRetry / withFallback entirely.\n      if (streamError !== undefined) {\n        throw new ProviderHttpError({\n          providerName,\n          status: 0,\n          message: streamError.message,\n        });\n      }\n      return {\n        text: collected.join(''),\n        usage,\n        finishReason,\n      };\n    },\n  };\n}\n\nfunction applyLlamaCppNodePreflight(\n  req: ProviderRequest,\n  capabilities: ProviderCapabilities,\n): ProviderRequest {\n  const retention = resolveReasoningRetention({\n    ...(req.reasoningRetention !== undefined ? { requested: req.reasoningRetention } : {}),\n    ...(capabilities.reasoningContract !== undefined\n      ? { contract: capabilities.reasoningContract }\n      : {}),\n  });\n  if (retention === 'pass-through-all') return req;\n  const filtered = applyReasoningPolicy({ messages: req.messages, retention });\n  if (filtered === req.messages) return req;\n  return { ...req, messages: filtered };\n}\n\nasync function* streamLlamaCppNode(\n  ensureModel: () => Promise<{ model: LlamaModelInstance }>,\n  sessionFactory: (model: LlamaModelInstance, system?: string) => Promise<LlamaSessionInstance>,\n  providerName: string,\n  options: LlamaCppNodeAdapterOptions,\n  req: ProviderRequest,\n): AsyncIterable<ProviderEvent> {\n  const { model } = await ensureModel();\n  const session = await sessionFactory(\n    model,\n    typeof req.systemMessage === 'string' ? req.systemMessage : undefined,\n  );\n  const prompt = renderPrompt(req);\n  const promptTokens = lengthOf(model.tokenize(prompt));\n  yield {\n    type: 'stream-start',\n    metadata: {\n      providerName,\n      modelId: options.modelPath,\n      createdAt: new Date().toISOString(),\n    },\n  };\n  let completionTokens = 0;\n  let errored = false;\n  let aborted = false;\n  try {\n    const streamOptions: { signal?: AbortSignal; maxTokens?: number; temperature?: number } = {};\n    if (req.signal !== undefined) streamOptions.signal = req.signal;\n    if (req.maxTokens !== undefined) streamOptions.maxTokens = req.maxTokens;\n    if (req.temperature !== undefined) streamOptions.temperature = req.temperature;\n    for await (const piece of session.promptStreamingResponse(prompt, streamOptions)) {\n      if (req.signal?.aborted) {\n        aborted = true;\n        break;\n      }\n      if (typeof piece !== 'string' || piece.length === 0) continue;\n      completionTokens += lengthOf(model.tokenize(piece));\n      yield { type: 'text-delta', delta: piece };\n      if (req.signal?.aborted) {\n        aborted = true;\n        break;\n      }\n    }\n  } catch (err) {\n    errored = true;\n    yield { type: 'error', error: { kind: 'unknown', message: (err as Error).message } };\n  } finally {\n    // Release the per-request context / KV cache (core-provider-08).\n    try {\n      session.dispose?.();\n    } catch {\n      // Disposal failures must never mask the stream outcome.\n    }\n  }\n  yield {\n    type: 'finish',\n    // PS-4: a mid-stream failure must not masquerade as a clean stop -\n    // partial text would be indistinguishable from success. PS-12: an\n    // aborted stream reports 'aborted', mirroring the HTTP adapters.\n    finishReason: errored ? 'error' : aborted ? 'aborted' : 'stop',\n    usage: {\n      promptTokens,\n      completionTokens,\n      totalTokens: promptTokens + completionTokens,\n    },\n  };\n}\n\nasync function resolveModel(options: LlamaCppNodeAdapterOptions): Promise<LlamaModelInstance> {\n  const llama = await loadLlamaModule(options.runtimeOverrides);\n  return llama.loadModel({\n    modelPath: options.modelPath,\n    ...(options.gpuLayers !== undefined ? { gpuLayers: options.gpuLayers } : {}),\n  });\n}\n\n/**\n * The REAL default session factory (PS-3): `model.createContext()` →\n * `context.getSequence()` → `new LlamaChatSession({ contextSequence })`\n * from the lazily-loaded peer, adapting its callback-streaming\n * `prompt(text, { onTextChunk })` to the adapter's\n * `promptStreamingResponse` AsyncIterable contract.\n */\nasync function defaultSessionFactory(\n  model: LlamaModelInstance,\n  system: string | undefined,\n  options: LlamaCppNodeAdapterOptions,\n): Promise<LlamaSessionInstance> {\n  const ChatSession = await loadLlamaChatSessionCtor(options.runtimeOverrides);\n  const context = await model.createContext(\n    options.contextSize !== undefined ? { contextSize: options.contextSize } : undefined,\n  );\n  const sequence = context.getSequence();\n  const session = new ChatSession({\n    contextSequence: sequence,\n    ...(system !== undefined ? { systemPrompt: system } : {}),\n  });\n  return {\n    promptStreamingResponse(prompt, streamOptions) {\n      return promptToIterable(session, prompt, streamOptions);\n    },\n    dispose() {\n      // Sequence first, then its owning context (core-provider-08).\n      try {\n        sequence.dispose?.();\n      } catch {\n        // fall through to the context\n      }\n      context.dispose?.();\n    },\n  };\n}\n\n/**\n * Bridge the peer's callback-streaming `prompt(...)` into the\n * AsyncIterable the adapter consumes. A rejection from `prompt`\n * rejects the pending/next iteration so the stream's error path\n * (PS-4) observes it.\n */\nfunction promptToIterable(\n  session: LlamaChatSessionPeer,\n  prompt: string,\n  streamOptions?: {\n    readonly signal?: AbortSignal;\n    readonly maxTokens?: number;\n    readonly temperature?: number;\n  },\n): AsyncIterable<string> {\n  const queue: string[] = [];\n  let done = false;\n  let failure: unknown;\n  let wake: (() => void) | null = null;\n  const notify = (): void => {\n    if (wake !== null) {\n      const w = wake;\n      wake = null;\n      w();\n    }\n  };\n  void session\n    .prompt(prompt, {\n      ...(streamOptions?.signal !== undefined ? { signal: streamOptions.signal } : {}),\n      ...(streamOptions?.maxTokens !== undefined ? { maxTokens: streamOptions.maxTokens } : {}),\n      ...(streamOptions?.temperature !== undefined\n        ? { temperature: streamOptions.temperature }\n        : {}),\n      onTextChunk: (chunk: string) => {\n        if (typeof chunk === 'string' && chunk.length > 0) queue.push(chunk);\n        notify();\n      },\n    })\n    .then(\n      () => {\n        done = true;\n        notify();\n      },\n      (err: unknown) => {\n        failure = err instanceof Error ? err : new Error(String(err));\n        done = true;\n        notify();\n      },\n    );\n  return {\n    [Symbol.asyncIterator](): AsyncIterator<string> {\n      return {\n        async next(): Promise<IteratorResult<string>> {\n          for (;;) {\n            const head = queue.shift();\n            if (head !== undefined) return { done: false, value: head };\n            if (failure !== undefined) throw failure;\n            if (done) return { done: true, value: undefined };\n            await new Promise<void>((resolve) => {\n              wake = resolve;\n            });\n          }\n        },\n      };\n    },\n  };\n}\n\nfunction renderPrompt(req: ProviderRequest): string {\n  const parts: string[] = [];\n  // `req.systemMessage` is deliberately NOT rendered here: the session\n  // factory already sets it as the chat-template `systemPrompt`, and\n  // rendering it again would show the model the system prompt twice\n  // (core-provider-08).\n  for (const msg of req.messages) {\n    const text =\n      typeof msg.content === 'string'\n        ? msg.content\n        : msg.content\n            .map((p) => (p.type === 'text' ? p.text : p.type === 'reasoning' ? p.text : ''))\n            .join('');\n    parts.push(`[${msg.role}] ${text}`);\n  }\n  return parts.join('\\n');\n}\n\nfunction basename(path: string): string {\n  const idx = Math.max(path.lastIndexOf('/'), path.lastIndexOf('\\\\'));\n  return idx === -1 ? path : path.slice(idx + 1);\n}\n\nfunction lengthOf(tokens: readonly number[] | Uint32Array | Uint8Array | null | undefined): number {\n  if (tokens === null || tokens === undefined) return 0;\n  return (tokens as { length?: number }).length ?? 0;\n}\n"],"mappings":";;;;;;;;;;;AAmCA,MAAaA,qCAAiE,OAAO,OAAO;CAC1F;CACA;CACA;CACD,CAAC;AA6CF,MAAMC,uBAA6C;CACjD,WAAW;CACX,aAAa;CACb,mBAAmB;CACnB,YAAY;CACZ,kBAAkB;CAClB,WAAW;CACX,eAAe;CACf,WAAW;CACX,mBAAmB;CACpB;;;;;;;;AASD,SAAgB,oBAAoB,SAA+C;CACjF,MAAM,eAAe,QAAQ,QAAQ,kBAAkB,SAAS,QAAQ,UAAU;CAClF,MAAMC,eAAqC;EACzC,GAAG;EACH,GAAG,QAAQ;EACZ;CACD,MAAM,qBAAqB,QAAQ,sBAAsB;CACzD,IAAIC,WAAiD;CACrD,IAAIC,YAA2D;CAC/D,MAAM,cAAc,YAAoD;AACtE,MAAI,aAAa,KAAM,QAAO;AAC9B,MAAI,cAAc,KAAM,QAAO;AAC/B,eAAa,YAAY;AAEvB,cAAW,EAAE,OADC,QAAQ,iBAAkB,MAAM,aAAa,QAAQ,EAC/C;AACpB,eAAY;AACZ,UAAO;MACL;AACJ,SAAO;;CAET,MAAM,iBACJ,QAAQ,kBACR,QAAQ,kBAAkB,mBACxB,OAA2B,WAAoB,sBAAsB,OAAO,QAAQ,QAAQ;AAChG,QAAO;EACL,MAAM;EACN,SAAS,QAAQ;EACjB;EACA;EACA,OAAO,KAAK;AACV,UAAO,mBACL,aACA,gBACA,cACA,SACA,2BAA2B,KAAK,aAAa,CAC9C;;EAEH,MAAM,SAAS,KAAK;GAClB,MAAM,SAAS,mBACb,aACA,gBACA,cACA,SACA,2BAA2B,KAAK,aAAa,CAC9C;GACD,MAAMC,YAAsB,EAAE;GAC9B,IAAIC,QAAe;IAAE,cAAc;IAAG,kBAAkB;IAAG,aAAa;IAAG;GAC3E,IAAIC,eAAiD;GACrD,IAAIC;AACJ,cAAW,MAAM,SAAS,QAAQ;AAChC,QAAI,MAAM,SAAS,aAAc,WAAU,KAAK,MAAM,MAAM;AAC5D,QAAI,MAAM,SAAS,QAAS,eAAc,MAAM;AAChD,QAAI,MAAM,SAAS,UAAU;AAC3B,aAAQ,MAAM;AACd,oBAAe,MAAM;;;AAMzB,OAAI,gBAAgB,OAClB,OAAM,IAAI,kBAAkB;IAC1B;IACA,QAAQ;IACR,SAAS,YAAY;IACtB,CAAC;AAEJ,UAAO;IACL,MAAM,UAAU,KAAK,GAAG;IACxB;IACA;IACD;;EAEJ;;AAGH,SAAS,2BACP,KACA,cACiB;CACjB,MAAM,YAAY,0BAA0B;EAC1C,GAAI,IAAI,uBAAuB,SAAY,EAAE,WAAW,IAAI,oBAAoB,GAAG,EAAE;EACrF,GAAI,aAAa,sBAAsB,SACnC,EAAE,UAAU,aAAa,mBAAmB,GAC5C,EAAE;EACP,CAAC;AACF,KAAI,cAAc,mBAAoB,QAAO;CAC7C,MAAM,WAAW,qBAAqB;EAAE,UAAU,IAAI;EAAU;EAAW,CAAC;AAC5E,KAAI,aAAa,IAAI,SAAU,QAAO;AACtC,QAAO;EAAE,GAAG;EAAK,UAAU;EAAU;;AAGvC,gBAAgB,mBACd,aACA,gBACA,cACA,SACA,KAC8B;CAC9B,MAAM,EAAE,UAAU,MAAM,aAAa;CACrC,MAAM,UAAU,MAAM,eACpB,OACA,OAAO,IAAI,kBAAkB,WAAW,IAAI,gBAAgB,OAC7D;CACD,MAAM,SAAS,aAAa,IAAI;CAChC,MAAM,eAAe,SAAS,MAAM,SAAS,OAAO,CAAC;AACrD,OAAM;EACJ,MAAM;EACN,UAAU;GACR;GACA,SAAS,QAAQ;GACjB,4BAAW,IAAI,MAAM,EAAC,aAAa;GACpC;EACF;CACD,IAAI,mBAAmB;CACvB,IAAI,UAAU;CACd,IAAI,UAAU;AACd,KAAI;EACF,MAAMC,gBAAoF,EAAE;AAC5F,MAAI,IAAI,WAAW,OAAW,eAAc,SAAS,IAAI;AACzD,MAAI,IAAI,cAAc,OAAW,eAAc,YAAY,IAAI;AAC/D,MAAI,IAAI,gBAAgB,OAAW,eAAc,cAAc,IAAI;AACnE,aAAW,MAAM,SAAS,QAAQ,wBAAwB,QAAQ,cAAc,EAAE;AAChF,OAAI,IAAI,QAAQ,SAAS;AACvB,cAAU;AACV;;AAEF,OAAI,OAAO,UAAU,YAAY,MAAM,WAAW,EAAG;AACrD,uBAAoB,SAAS,MAAM,SAAS,MAAM,CAAC;AACnD,SAAM;IAAE,MAAM;IAAc,OAAO;IAAO;AAC1C,OAAI,IAAI,QAAQ,SAAS;AACvB,cAAU;AACV;;;UAGG,KAAK;AACZ,YAAU;AACV,QAAM;GAAE,MAAM;GAAS,OAAO;IAAE,MAAM;IAAW,SAAU,IAAc;IAAS;GAAE;WAC5E;AAER,MAAI;AACF,WAAQ,WAAW;UACb;;AAIV,OAAM;EACJ,MAAM;EAIN,cAAc,UAAU,UAAU,UAAU,YAAY;EACxD,OAAO;GACL;GACA;GACA,aAAa,eAAe;GAC7B;EACF;;AAGH,eAAe,aAAa,SAAkE;AAE5F,SADc,MAAM,gBAAgB,QAAQ,iBAAiB,EAChD,UAAU;EACrB,WAAW,QAAQ;EACnB,GAAI,QAAQ,cAAc,SAAY,EAAE,WAAW,QAAQ,WAAW,GAAG,EAAE;EAC5E,CAAC;;;;;;;;;AAUJ,eAAe,sBACb,OACA,QACA,SAC+B;CAC/B,MAAM,cAAc,MAAM,yBAAyB,QAAQ,iBAAiB;CAC5E,MAAM,UAAU,MAAM,MAAM,cAC1B,QAAQ,gBAAgB,SAAY,EAAE,aAAa,QAAQ,aAAa,GAAG,OAC5E;CACD,MAAM,WAAW,QAAQ,aAAa;CACtC,MAAM,UAAU,IAAI,YAAY;EAC9B,iBAAiB;EACjB,GAAI,WAAW,SAAY,EAAE,cAAc,QAAQ,GAAG,EAAE;EACzD,CAAC;AACF,QAAO;EACL,wBAAwB,QAAQ,eAAe;AAC7C,UAAO,iBAAiB,SAAS,QAAQ,cAAc;;EAEzD,UAAU;AAER,OAAI;AACF,aAAS,WAAW;WACd;AAGR,WAAQ,WAAW;;EAEtB;;;;;;;;AASH,SAAS,iBACP,SACA,QACA,eAKuB;CACvB,MAAMC,QAAkB,EAAE;CAC1B,IAAI,OAAO;CACX,IAAIC;CACJ,IAAIC,OAA4B;CAChC,MAAM,eAAqB;AACzB,MAAI,SAAS,MAAM;GACjB,MAAM,IAAI;AACV,UAAO;AACP,MAAG;;;AAGP,CAAK,QACF,OAAO,QAAQ;EACd,GAAI,eAAe,WAAW,SAAY,EAAE,QAAQ,cAAc,QAAQ,GAAG,EAAE;EAC/E,GAAI,eAAe,cAAc,SAAY,EAAE,WAAW,cAAc,WAAW,GAAG,EAAE;EACxF,GAAI,eAAe,gBAAgB,SAC/B,EAAE,aAAa,cAAc,aAAa,GAC1C,EAAE;EACN,cAAc,UAAkB;AAC9B,OAAI,OAAO,UAAU,YAAY,MAAM,SAAS,EAAG,OAAM,KAAK,MAAM;AACpE,WAAQ;;EAEX,CAAC,CACD,WACO;AACJ,SAAO;AACP,UAAQ;KAET,QAAiB;AAChB,YAAU,eAAe,QAAQ,MAAM,IAAI,MAAM,OAAO,IAAI,CAAC;AAC7D,SAAO;AACP,UAAQ;GAEX;AACH,QAAO,EACL,CAAC,OAAO,iBAAwC;AAC9C,SAAO,EACL,MAAM,OAAwC;AAC5C,YAAS;IACP,MAAM,OAAO,MAAM,OAAO;AAC1B,QAAI,SAAS,OAAW,QAAO;KAAE,MAAM;KAAO,OAAO;KAAM;AAC3D,QAAI,YAAY,OAAW,OAAM;AACjC,QAAI,KAAM,QAAO;KAAE,MAAM;KAAM,OAAO;KAAW;AACjD,UAAM,IAAI,SAAe,YAAY;AACnC,YAAO;MACP;;KAGP;IAEJ;;AAGH,SAAS,aAAa,KAA8B;CAClD,MAAMC,QAAkB,EAAE;AAK1B,MAAK,MAAM,OAAO,IAAI,UAAU;EAC9B,MAAM,OACJ,OAAO,IAAI,YAAY,WACnB,IAAI,UACJ,IAAI,QACD,KAAK,MAAO,EAAE,SAAS,SAAS,EAAE,OAAO,EAAE,SAAS,cAAc,EAAE,OAAO,GAAI,CAC/E,KAAK,GAAG;AACjB,QAAM,KAAK,IAAI,IAAI,KAAK,IAAI,OAAO;;AAErC,QAAO,MAAM,KAAK,KAAK;;AAGzB,SAAS,SAAS,MAAsB;CACtC,MAAM,MAAM,KAAK,IAAI,KAAK,YAAY,IAAI,EAAE,KAAK,YAAY,KAAK,CAAC;AACnE,QAAO,QAAQ,KAAK,OAAO,KAAK,MAAM,MAAM,EAAE;;AAGhD,SAAS,SAAS,QAAiF;AACjG,KAAI,WAAW,QAAQ,WAAW,OAAW,QAAO;AACpD,QAAQ,OAA+B,UAAU"}
+{"version":3,"file":"adapter.js","names":["LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY: ReadonlyArray<Sensitivity>","DEFAULT_CAPABILITIES: ProviderCapabilities","capabilities: ProviderCapabilities","resolved: { model: LlamaModelInstance } | null","resolving: Promise<{ model: LlamaModelInstance }> | null","collected: string[]","usage: Usage","finishReason: ProviderResponse['finishReason']","streamError: { message: string } | undefined","prompt: string","pieces: string[]","streamOptions: { signal?: AbortSignal; maxTokens?: number; temperature?: number }","cached: LlamaSessionInstance | null","queueTail: Promise<void>","unlock!: () => void","turns: LlamaChatHistoryItem[]","queue: string[]","failure: unknown","wake: (() => void) | null","parts: string[]"],"sources":["../src/adapter.ts"],"sourcesContent":["/**\n * `llamaCppNodeAdapter` - in-process GGUF adapter built on\n * `node-llama-cpp@^3.5`. The adapter declares `trust: 'loopback'`\n * permanently because the model lives in the same trust boundary as\n * the host process.\n *\n * @packageDocumentation\n */\n\nimport type {\n  Provider,\n  ProviderCapabilities,\n  ProviderEvent,\n  ProviderRequest,\n  ProviderResponse,\n  Sensitivity,\n  Usage,\n} from '@graphorin/core';\nimport { ProviderHttpError } from '@graphorin/provider/errors';\nimport { applyReasoningPolicy, resolveReasoningRetention } from '@graphorin/provider/reasoning';\nimport {\n  type LlamaChatHistoryItem,\n  type LlamaChatSessionPeer,\n  type LlamaCppNodeRuntimeOverrides,\n  type LlamaModelInstance,\n  type LlamaSessionInstance,\n  loadLlamaChatSessionCtor,\n  loadLlamaModule,\n} from './runtime.js';\n\n/**\n * Default sensitivity envelope for the in-process adapter - same as\n * the loopback HTTP path.\n *\n * @stable\n */\nexport const LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY: ReadonlyArray<Sensitivity> = Object.freeze([\n  'public',\n  'internal',\n  'secret',\n]);\n\n/**\n * Options accepted by {@link llamaCppNodeAdapter}.\n *\n * @stable\n */\nexport interface LlamaCppNodeAdapterOptions {\n  /** Filesystem path to the `.gguf` model file. */\n  readonly modelPath: string;\n  /** Number of layers to offload to the GPU. Default `'auto'`. */\n  readonly gpuLayers?: number | 'auto';\n  /** Optional context-window override. */\n  readonly contextSize?: number;\n  /** Provider name attached to spans / log lines. */\n  readonly name?: string;\n  /** Capability declaration. Merged on top of the defaults table. */\n  readonly capabilities?: Partial<ProviderCapabilities>;\n  /** Sensitivity override (defaults to the loopback envelope). */\n  readonly acceptsSensitivity?: ReadonlyArray<Sensitivity>;\n  /**\n   * Test-only runtime override. When unset the adapter loads\n   * `node-llama-cpp` lazily on first call.\n   */\n  readonly runtimeOverrides?: LlamaCppNodeRuntimeOverrides;\n  /**\n   * Optional `model` override that short-circuits the\n   * `loadLlamaModule(...).loadModel(...)` flow. Tests pass a fixture\n   * shaped instance.\n   */\n  readonly modelOverride?: LlamaModelInstance;\n  /**\n   * Optional session factory override. When unset, the adapter builds a\n   * real session from the peer (PS-3): `model.createContext()` →\n   * `new LlamaChatSession({ contextSequence })`, streaming through\n   * `prompt(text, { onTextChunk })`. Overrides\n   * (`runtimeOverrides.createSession` or this option) keep the test\n   * seam.\n   */\n  readonly sessionFactory?: (\n    model: LlamaModelInstance,\n    system?: string,\n  ) => Promise<LlamaSessionInstance>;\n  /**\n   * W-096: reuse ONE session (context + KV cache) across requests\n   * instead of creating and disposing a fresh one per call - an agent\n   * loop then avoids re-prefilling the growing transcript on every\n   * step. Requests serialise through a promise mutex (a llama context\n   * sequence is single-threaded), and the chat history re-syncs via\n   * `setChatHistory` before each prompt. Strictly opt-in: the default\n   * per-request lifecycle stays memory-safe and concurrency-safe; the\n   * cached session also skips per-request disposal (it lives until the\n   * process / instance is released). Sessions WITHOUT `setChatHistory`\n   * cannot re-sync and silently degrade to per-request behaviour.\n   */\n  readonly persistentSession?: boolean;\n}\n\nconst DEFAULT_CAPABILITIES: ProviderCapabilities = {\n  streaming: true,\n  toolCalling: false,\n  parallelToolCalls: false,\n  multimodal: false,\n  structuredOutput: false,\n  reasoning: false,\n  contextWindow: 8_192,\n  maxOutput: 4_096,\n  reasoningContract: 'optional',\n};\n\n/**\n * Build a Graphorin {@link Provider} backed by an in-process GGUF\n * model. The first call lazily loads the `node-llama-cpp` peer + the\n * model file; subsequent calls reuse the cached instances.\n *\n * @stable\n */\nexport function llamaCppNodeAdapter(options: LlamaCppNodeAdapterOptions): Provider {\n  const providerName = options.name ?? `llama-cpp-node-${basename(options.modelPath)}`;\n  const capabilities: ProviderCapabilities = {\n    ...DEFAULT_CAPABILITIES,\n    ...options.capabilities,\n  };\n  const acceptsSensitivity = options.acceptsSensitivity ?? LLAMA_CPP_NODE_ACCEPTS_SENSITIVITY;\n  let resolved: { model: LlamaModelInstance } | null = null;\n  let resolving: Promise<{ model: LlamaModelInstance }> | null = null;\n  const ensureModel = async (): Promise<{ model: LlamaModelInstance }> => {\n    if (resolved !== null) return resolved;\n    if (resolving !== null) return resolving;\n    resolving = (async () => {\n      const model = options.modelOverride ?? (await resolveModel(options));\n      resolved = { model };\n      resolving = null;\n      return resolved;\n    })();\n    return resolving;\n  };\n  const sessionFactory =\n    options.sessionFactory ??\n    options.runtimeOverrides?.createSession ??\n    ((model: LlamaModelInstance, system?: string) => defaultSessionFactory(model, system, options));\n  const acquireSession = buildSessionManager(options, sessionFactory);\n  return {\n    name: providerName,\n    modelId: options.modelPath,\n    capabilities,\n    acceptsSensitivity,\n    stream(req) {\n      return streamLlamaCppNode(\n        ensureModel,\n        acquireSession,\n        providerName,\n        options,\n        applyLlamaCppNodePreflight(req, capabilities),\n      );\n    },\n    async generate(req) {\n      const events = streamLlamaCppNode(\n        ensureModel,\n        acquireSession,\n        providerName,\n        options,\n        applyLlamaCppNodePreflight(req, capabilities),\n      );\n      const collected: string[] = [];\n      let usage: Usage = { promptTokens: 0, completionTokens: 0, totalTokens: 0 };\n      let finishReason: ProviderResponse['finishReason'] = 'stop';\n      let streamError: { message: string } | undefined;\n      for await (const event of events) {\n        if (event.type === 'text-delta') collected.push(event.delta);\n        if (event.type === 'error') streamError = event.error;\n        if (event.type === 'finish') {\n          usage = event.usage;\n          finishReason = event.finishReason;\n        }\n      }\n      // PS-4: a swallowed mid-stream error returned truncated text\n      // indistinguishable from success - and a never-throwing\n      // generate() bypasses withRetry / withFallback entirely.\n      if (streamError !== undefined) {\n        throw new ProviderHttpError({\n          providerName,\n          status: 0,\n          message: streamError.message,\n        });\n      }\n      return {\n        text: collected.join(''),\n        usage,\n        finishReason,\n      };\n    },\n  };\n}\n\nfunction applyLlamaCppNodePreflight(\n  req: ProviderRequest,\n  capabilities: ProviderCapabilities,\n): ProviderRequest {\n  const retention = resolveReasoningRetention({\n    ...(req.reasoningRetention !== undefined ? { requested: req.reasoningRetention } : {}),\n    ...(capabilities.reasoningContract !== undefined\n      ? { contract: capabilities.reasoningContract }\n      : {}),\n  });\n  if (retention === 'pass-through-all') return req;\n  const filtered = applyReasoningPolicy({ messages: req.messages, retention });\n  if (filtered === req.messages) return req;\n  return { ...req, messages: filtered };\n}\n\nasync function* streamLlamaCppNode(\n  ensureModel: () => Promise<{ model: LlamaModelInstance }>,\n  acquireSession: SessionManager,\n  providerName: string,\n  options: LlamaCppNodeAdapterOptions,\n  req: ProviderRequest,\n): AsyncIterable<ProviderEvent> {\n  const { model } = await ensureModel();\n  const lease = await acquireSession(\n    model,\n    typeof req.systemMessage === 'string' ? req.systemMessage : undefined,\n  );\n  const session = lease.session;\n  // W-096: feed the transcript as REAL chat history when the session\n  // supports it (node-llama-cpp v3 `setChatHistory`) and prompt ONLY\n  // the last user turn - the pre-fix pseudo-prompt double-templated the\n  // whole conversation inside one user turn, degrading chat-tuned GGUF\n  // models. Sessions without the setter (older fixtures, custom\n  // factories) keep the legacy renderPrompt path unchanged.\n  const historic = typeof session.setChatHistory === 'function';\n  let prompt: string;\n  if (historic) {\n    const { priorTurns, lastUserText } = buildChatHistory(req);\n    session.setChatHistory?.(priorTurns);\n    prompt = lastUserText;\n  } else {\n    prompt = renderPrompt(req);\n  }\n  // Prompt-token accounting stays the full-transcript proxy in both\n  // modes so the figure is comparable regardless of the wire path.\n  const promptTokens = lengthOf(model.tokenize(renderPrompt(req)));\n  yield {\n    type: 'stream-start',\n    metadata: {\n      providerName,\n      modelId: options.modelPath,\n      createdAt: new Date().toISOString(),\n    },\n  };\n  const pieces: string[] = [];\n  let errored = false;\n  let aborted = false;\n  try {\n    const streamOptions: { signal?: AbortSignal; maxTokens?: number; temperature?: number } = {};\n    if (req.signal !== undefined) streamOptions.signal = req.signal;\n    if (req.maxTokens !== undefined) streamOptions.maxTokens = req.maxTokens;\n    if (req.temperature !== undefined) streamOptions.temperature = req.temperature;\n    for await (const piece of session.promptStreamingResponse(prompt, streamOptions)) {\n      if (req.signal?.aborted) {\n        aborted = true;\n        break;\n      }\n      if (typeof piece !== 'string' || piece.length === 0) continue;\n      pieces.push(piece);\n      yield { type: 'text-delta', delta: piece };\n      if (req.signal?.aborted) {\n        aborted = true;\n        break;\n      }\n    }\n  } catch (err) {\n    errored = true;\n    yield { type: 'error', error: { kind: 'unknown', message: (err as Error).message } };\n  } finally {\n    lease.release();\n  }\n  // W-096: tokenize the ASSEMBLED response once - per-chunk tokenization\n  // over-counted at every chunk boundary (BPE merges span chunk seams).\n  const completionTokens = lengthOf(model.tokenize(pieces.join('')));\n  yield {\n    type: 'finish',\n    // PS-4: a mid-stream failure must not masquerade as a clean stop -\n    // partial text would be indistinguishable from success. PS-12: an\n    // aborted stream reports 'aborted', mirroring the HTTP adapters.\n    finishReason: errored ? 'error' : aborted ? 'aborted' : 'stop',\n    usage: {\n      promptTokens,\n      completionTokens,\n      totalTokens: promptTokens + completionTokens,\n    },\n  };\n}\n\n/** Lease over a session: per-request leases dispose on release. */\ninterface SessionLease {\n  readonly session: LlamaSessionInstance;\n  release(): void;\n}\n\ntype SessionManager = (model: LlamaModelInstance, system?: string) => Promise<SessionLease>;\n\n/**\n * W-096: per-request sessions by default (create -> use -> dispose, the\n * memory-safe core-provider-08 lifecycle); with `persistentSession:\n * true` ONE cached session serialised by a promise mutex (a llama\n * context sequence is single-threaded) whose history re-syncs via\n * `setChatHistory` on every request. A cached session that turns out\n * not to support `setChatHistory` degrades to per-request behaviour -\n * without the setter its history cannot be made consistent.\n */\nfunction buildSessionManager(\n  options: LlamaCppNodeAdapterOptions,\n  sessionFactory: (model: LlamaModelInstance, system?: string) => Promise<LlamaSessionInstance>,\n): SessionManager {\n  if (options.persistentSession !== true) {\n    return async (model, system) => {\n      const session = await sessionFactory(model, system);\n      return {\n        session,\n        release: () => {\n          // Release the per-request context / KV cache (core-provider-08).\n          try {\n            session.dispose?.();\n          } catch {\n            // Disposal failures must never mask the stream outcome.\n          }\n        },\n      };\n    };\n  }\n  let cached: LlamaSessionInstance | null = null;\n  let queueTail: Promise<void> = Promise.resolve();\n  return async (model, system) => {\n    let unlock!: () => void;\n    const gate = new Promise<void>((resolve) => {\n      unlock = resolve;\n    });\n    const priorTail = queueTail;\n    queueTail = queueTail.then(() => gate);\n    await priorTail;\n    if (cached === null) cached = await sessionFactory(model, system);\n    const session = cached;\n    if (typeof session.setChatHistory !== 'function') {\n      // Cannot re-sync history across requests - fall back to the\n      // per-request lifecycle for THIS session and try again next call.\n      cached = null;\n      return {\n        session,\n        release: () => {\n          try {\n            session.dispose?.();\n          } catch {\n            // best-effort\n          }\n          unlock();\n        },\n      };\n    }\n    return { session, release: () => unlock() };\n  };\n}\n\n/**\n * W-096: split the request into node-llama-cpp chat-history turns plus\n * the trailing user text (the prompt). The system message is included\n * as a history item because `setChatHistory` REPLACES the session's\n * whole history - including the slot the constructor `systemPrompt`\n * seeded - so relying on the constructor alone would drop it. Tool\n * roles cannot occur (`toolCalling: false`); a defensive skip keeps a\n * stray one from corrupting the template.\n */\nfunction buildChatHistory(req: ProviderRequest): {\n  readonly priorTurns: ReadonlyArray<LlamaChatHistoryItem>;\n  readonly lastUserText: string;\n} {\n  const turns: LlamaChatHistoryItem[] = [];\n  if (typeof req.systemMessage === 'string' && req.systemMessage.length > 0) {\n    turns.push({ type: 'system', text: req.systemMessage });\n  }\n  for (const msg of req.messages) {\n    const text = textOf(msg);\n    if (msg.role === 'user') turns.push({ type: 'user', text });\n    else if (msg.role === 'assistant') turns.push({ type: 'model', response: [text] });\n    else if (msg.role === 'system') turns.push({ type: 'system', text });\n  }\n  const last = turns[turns.length - 1];\n  if (last !== undefined && last.type === 'user') {\n    turns.pop();\n    return { priorTurns: turns, lastUserText: last.text };\n  }\n  // No trailing user turn (continuation) - full history, empty prompt.\n  return { priorTurns: turns, lastUserText: '' };\n}\n\nfunction textOf(msg: ProviderRequest['messages'][number]): string {\n  return typeof msg.content === 'string'\n    ? msg.content\n    : msg.content\n        .map((p) => (p.type === 'text' ? p.text : p.type === 'reasoning' ? p.text : ''))\n        .join('');\n}\n\nasync function resolveModel(options: LlamaCppNodeAdapterOptions): Promise<LlamaModelInstance> {\n  const llama = await loadLlamaModule(options.runtimeOverrides);\n  return llama.loadModel({\n    modelPath: options.modelPath,\n    ...(options.gpuLayers !== undefined ? { gpuLayers: options.gpuLayers } : {}),\n  });\n}\n\n/**\n * The REAL default session factory (PS-3): `model.createContext()` →\n * `context.getSequence()` → `new LlamaChatSession({ contextSequence })`\n * from the lazily-loaded peer, adapting its callback-streaming\n * `prompt(text, { onTextChunk })` to the adapter's\n * `promptStreamingResponse` AsyncIterable contract.\n */\nasync function defaultSessionFactory(\n  model: LlamaModelInstance,\n  system: string | undefined,\n  options: LlamaCppNodeAdapterOptions,\n): Promise<LlamaSessionInstance> {\n  const ChatSession = await loadLlamaChatSessionCtor(options.runtimeOverrides);\n  const context = await model.createContext(\n    options.contextSize !== undefined ? { contextSize: options.contextSize } : undefined,\n  );\n  const sequence = context.getSequence();\n  const session = new ChatSession({\n    contextSequence: sequence,\n    ...(system !== undefined ? { systemPrompt: system } : {}),\n  });\n  return {\n    promptStreamingResponse(prompt, streamOptions) {\n      return promptToIterable(session, prompt, streamOptions);\n    },\n    dispose() {\n      // Sequence first, then its owning context (core-provider-08).\n      try {\n        sequence.dispose?.();\n      } catch {\n        // fall through to the context\n      }\n      context.dispose?.();\n    },\n  };\n}\n\n/**\n * Bridge the peer's callback-streaming `prompt(...)` into the\n * AsyncIterable the adapter consumes. A rejection from `prompt`\n * rejects the pending/next iteration so the stream's error path\n * (PS-4) observes it.\n */\nfunction promptToIterable(\n  session: LlamaChatSessionPeer,\n  prompt: string,\n  streamOptions?: {\n    readonly signal?: AbortSignal;\n    readonly maxTokens?: number;\n    readonly temperature?: number;\n  },\n): AsyncIterable<string> {\n  const queue: string[] = [];\n  let done = false;\n  let failure: unknown;\n  let wake: (() => void) | null = null;\n  const notify = (): void => {\n    if (wake !== null) {\n      const w = wake;\n      wake = null;\n      w();\n    }\n  };\n  void session\n    .prompt(prompt, {\n      ...(streamOptions?.signal !== undefined ? { signal: streamOptions.signal } : {}),\n      ...(streamOptions?.maxTokens !== undefined ? { maxTokens: streamOptions.maxTokens } : {}),\n      ...(streamOptions?.temperature !== undefined\n        ? { temperature: streamOptions.temperature }\n        : {}),\n      onTextChunk: (chunk: string) => {\n        if (typeof chunk === 'string' && chunk.length > 0) queue.push(chunk);\n        notify();\n      },\n    })\n    .then(\n      () => {\n        done = true;\n        notify();\n      },\n      (err: unknown) => {\n        failure = err instanceof Error ? err : new Error(String(err));\n        done = true;\n        notify();\n      },\n    );\n  return {\n    [Symbol.asyncIterator](): AsyncIterator<string> {\n      return {\n        async next(): Promise<IteratorResult<string>> {\n          for (;;) {\n            const head = queue.shift();\n            if (head !== undefined) return { done: false, value: head };\n            if (failure !== undefined) throw failure;\n            if (done) return { done: true, value: undefined };\n            await new Promise<void>((resolve) => {\n              wake = resolve;\n            });\n          }\n        },\n      };\n    },\n  };\n}\n\nfunction renderPrompt(req: ProviderRequest): string {\n  const parts: string[] = [];\n  // `req.systemMessage` is deliberately NOT rendered here: the session\n  // factory already sets it as the chat-template `systemPrompt`, and\n  // rendering it again would show the model the system prompt twice\n  // (core-provider-08).\n  for (const msg of req.messages) {\n    const text =\n      typeof msg.content === 'string'\n        ? msg.content\n        : msg.content\n            .map((p) => (p.type === 'text' ? p.text : p.type === 'reasoning' ? p.text : ''))\n            .join('');\n    parts.push(`[${msg.role}] ${text}`);\n  }\n  return parts.join('\\n');\n}\n\nfunction basename(path: string): string {\n  const idx = Math.max(path.lastIndexOf('/'), path.lastIndexOf('\\\\'));\n  return idx === -1 ? path : path.slice(idx + 1);\n}\n\nfunction lengthOf(tokens: readonly number[] | Uint32Array | Uint8Array | null | undefined): number {\n  if (tokens === null || tokens === undefined) return 0;\n  return (tokens as { length?: number }).length ?? 0;\n}\n"],"mappings":";;;;;;;;;;;AAoCA,MAAaA,qCAAiE,OAAO,OAAO;CAC1F;CACA;CACA;CACD,CAAC;AA0DF,MAAMC,uBAA6C;CACjD,WAAW;CACX,aAAa;CACb,mBAAmB;CACnB,YAAY;CACZ,kBAAkB;CAClB,WAAW;CACX,eAAe;CACf,WAAW;CACX,mBAAmB;CACpB;;;;;;;;AASD,SAAgB,oBAAoB,SAA+C;CACjF,MAAM,eAAe,QAAQ,QAAQ,kBAAkB,SAAS,QAAQ,UAAU;CAClF,MAAMC,eAAqC;EACzC,GAAG;EACH,GAAG,QAAQ;EACZ;CACD,MAAM,qBAAqB,QAAQ,sBAAsB;CACzD,IAAIC,WAAiD;CACrD,IAAIC,YAA2D;CAC/D,MAAM,cAAc,YAAoD;AACtE,MAAI,aAAa,KAAM,QAAO;AAC9B,MAAI,cAAc,KAAM,QAAO;AAC/B,eAAa,YAAY;AAEvB,cAAW,EAAE,OADC,QAAQ,iBAAkB,MAAM,aAAa,QAAQ,EAC/C;AACpB,eAAY;AACZ,UAAO;MACL;AACJ,SAAO;;CAMT,MAAM,iBAAiB,oBAAoB,SAHzC,QAAQ,kBACR,QAAQ,kBAAkB,mBACxB,OAA2B,WAAoB,sBAAsB,OAAO,QAAQ,QAAQ,EAC7B;AACnE,QAAO;EACL,MAAM;EACN,SAAS,QAAQ;EACjB;EACA;EACA,OAAO,KAAK;AACV,UAAO,mBACL,aACA,gBACA,cACA,SACA,2BAA2B,KAAK,aAAa,CAC9C;;EAEH,MAAM,SAAS,KAAK;GAClB,MAAM,SAAS,mBACb,aACA,gBACA,cACA,SACA,2BAA2B,KAAK,aAAa,CAC9C;GACD,MAAMC,YAAsB,EAAE;GAC9B,IAAIC,QAAe;IAAE,cAAc;IAAG,kBAAkB;IAAG,aAAa;IAAG;GAC3E,IAAIC,eAAiD;GACrD,IAAIC;AACJ,cAAW,MAAM,SAAS,QAAQ;AAChC,QAAI,MAAM,SAAS,aAAc,WAAU,KAAK,MAAM,MAAM;AAC5D,QAAI,MAAM,SAAS,QAAS,eAAc,MAAM;AAChD,QAAI,MAAM,SAAS,UAAU;AAC3B,aAAQ,MAAM;AACd,oBAAe,MAAM;;;AAMzB,OAAI,gBAAgB,OAClB,OAAM,IAAI,kBAAkB;IAC1B;IACA,QAAQ;IACR,SAAS,YAAY;IACtB,CAAC;AAEJ,UAAO;IACL,MAAM,UAAU,KAAK,GAAG;IACxB;IACA;IACD;;EAEJ;;AAGH,SAAS,2BACP,KACA,cACiB;CACjB,MAAM,YAAY,0BAA0B;EAC1C,GAAI,IAAI,uBAAuB,SAAY,EAAE,WAAW,IAAI,oBAAoB,GAAG,EAAE;EACrF,GAAI,aAAa,sBAAsB,SACnC,EAAE,UAAU,aAAa,mBAAmB,GAC5C,EAAE;EACP,CAAC;AACF,KAAI,cAAc,mBAAoB,QAAO;CAC7C,MAAM,WAAW,qBAAqB;EAAE,UAAU,IAAI;EAAU;EAAW,CAAC;AAC5E,KAAI,aAAa,IAAI,SAAU,QAAO;AACtC,QAAO;EAAE,GAAG;EAAK,UAAU;EAAU;;AAGvC,gBAAgB,mBACd,aACA,gBACA,cACA,SACA,KAC8B;CAC9B,MAAM,EAAE,UAAU,MAAM,aAAa;CACrC,MAAM,QAAQ,MAAM,eAClB,OACA,OAAO,IAAI,kBAAkB,WAAW,IAAI,gBAAgB,OAC7D;CACD,MAAM,UAAU,MAAM;CAOtB,MAAM,WAAW,OAAO,QAAQ,mBAAmB;CACnD,IAAIC;AACJ,KAAI,UAAU;EACZ,MAAM,EAAE,YAAY,iBAAiB,iBAAiB,IAAI;AAC1D,UAAQ,iBAAiB,WAAW;AACpC,WAAS;OAET,UAAS,aAAa,IAAI;CAI5B,MAAM,eAAe,SAAS,MAAM,SAAS,aAAa,IAAI,CAAC,CAAC;AAChE,OAAM;EACJ,MAAM;EACN,UAAU;GACR;GACA,SAAS,QAAQ;GACjB,4BAAW,IAAI,MAAM,EAAC,aAAa;GACpC;EACF;CACD,MAAMC,SAAmB,EAAE;CAC3B,IAAI,UAAU;CACd,IAAI,UAAU;AACd,KAAI;EACF,MAAMC,gBAAoF,EAAE;AAC5F,MAAI,IAAI,WAAW,OAAW,eAAc,SAAS,IAAI;AACzD,MAAI,IAAI,cAAc,OAAW,eAAc,YAAY,IAAI;AAC/D,MAAI,IAAI,gBAAgB,OAAW,eAAc,cAAc,IAAI;AACnE,aAAW,MAAM,SAAS,QAAQ,wBAAwB,QAAQ,cAAc,EAAE;AAChF,OAAI,IAAI,QAAQ,SAAS;AACvB,cAAU;AACV;;AAEF,OAAI,OAAO,UAAU,YAAY,MAAM,WAAW,EAAG;AACrD,UAAO,KAAK,MAAM;AAClB,SAAM;IAAE,MAAM;IAAc,OAAO;IAAO;AAC1C,OAAI,IAAI,QAAQ,SAAS;AACvB,cAAU;AACV;;;UAGG,KAAK;AACZ,YAAU;AACV,QAAM;GAAE,MAAM;GAAS,OAAO;IAAE,MAAM;IAAW,SAAU,IAAc;IAAS;GAAE;WAC5E;AACR,QAAM,SAAS;;CAIjB,MAAM,mBAAmB,SAAS,MAAM,SAAS,OAAO,KAAK,GAAG,CAAC,CAAC;AAClE,OAAM;EACJ,MAAM;EAIN,cAAc,UAAU,UAAU,UAAU,YAAY;EACxD,OAAO;GACL;GACA;GACA,aAAa,eAAe;GAC7B;EACF;;;;;;;;;;;AAoBH,SAAS,oBACP,SACA,gBACgB;AAChB,KAAI,QAAQ,sBAAsB,KAChC,QAAO,OAAO,OAAO,WAAW;EAC9B,MAAM,UAAU,MAAM,eAAe,OAAO,OAAO;AACnD,SAAO;GACL;GACA,eAAe;AAEb,QAAI;AACF,aAAQ,WAAW;YACb;;GAIX;;CAGL,IAAIC,SAAsC;CAC1C,IAAIC,YAA2B,QAAQ,SAAS;AAChD,QAAO,OAAO,OAAO,WAAW;EAC9B,IAAIC;EACJ,MAAM,OAAO,IAAI,SAAe,YAAY;AAC1C,YAAS;IACT;EACF,MAAM,YAAY;AAClB,cAAY,UAAU,WAAW,KAAK;AACtC,QAAM;AACN,MAAI,WAAW,KAAM,UAAS,MAAM,eAAe,OAAO,OAAO;EACjE,MAAM,UAAU;AAChB,MAAI,OAAO,QAAQ,mBAAmB,YAAY;AAGhD,YAAS;AACT,UAAO;IACL;IACA,eAAe;AACb,SAAI;AACF,cAAQ,WAAW;aACb;AAGR,aAAQ;;IAEX;;AAEH,SAAO;GAAE;GAAS,eAAe,QAAQ;GAAE;;;;;;;;;;;;AAa/C,SAAS,iBAAiB,KAGxB;CACA,MAAMC,QAAgC,EAAE;AACxC,KAAI,OAAO,IAAI,kBAAkB,YAAY,IAAI,cAAc,SAAS,EACtE,OAAM,KAAK;EAAE,MAAM;EAAU,MAAM,IAAI;EAAe,CAAC;AAEzD,MAAK,MAAM,OAAO,IAAI,UAAU;EAC9B,MAAM,OAAO,OAAO,IAAI;AACxB,MAAI,IAAI,SAAS,OAAQ,OAAM,KAAK;GAAE,MAAM;GAAQ;GAAM,CAAC;WAClD,IAAI,SAAS,YAAa,OAAM,KAAK;GAAE,MAAM;GAAS,UAAU,CAAC,KAAK;GAAE,CAAC;WACzE,IAAI,SAAS,SAAU,OAAM,KAAK;GAAE,MAAM;GAAU;GAAM,CAAC;;CAEtE,MAAM,OAAO,MAAM,MAAM,SAAS;AAClC,KAAI,SAAS,UAAa,KAAK,SAAS,QAAQ;AAC9C,QAAM,KAAK;AACX,SAAO;GAAE,YAAY;GAAO,cAAc,KAAK;GAAM;;AAGvD,QAAO;EAAE,YAAY;EAAO,cAAc;EAAI;;AAGhD,SAAS,OAAO,KAAkD;AAChE,QAAO,OAAO,IAAI,YAAY,WAC1B,IAAI,UACJ,IAAI,QACD,KAAK,MAAO,EAAE,SAAS,SAAS,EAAE,OAAO,EAAE,SAAS,cAAc,EAAE,OAAO,GAAI,CAC/E,KAAK,GAAG;;AAGjB,eAAe,aAAa,SAAkE;AAE5F,SADc,MAAM,gBAAgB,QAAQ,iBAAiB,EAChD,UAAU;EACrB,WAAW,QAAQ;EACnB,GAAI,QAAQ,cAAc,SAAY,EAAE,WAAW,QAAQ,WAAW,GAAG,EAAE;EAC5E,CAAC;;;;;;;;;AAUJ,eAAe,sBACb,OACA,QACA,SAC+B;CAC/B,MAAM,cAAc,MAAM,yBAAyB,QAAQ,iBAAiB;CAC5E,MAAM,UAAU,MAAM,MAAM,cAC1B,QAAQ,gBAAgB,SAAY,EAAE,aAAa,QAAQ,aAAa,GAAG,OAC5E;CACD,MAAM,WAAW,QAAQ,aAAa;CACtC,MAAM,UAAU,IAAI,YAAY;EAC9B,iBAAiB;EACjB,GAAI,WAAW,SAAY,EAAE,cAAc,QAAQ,GAAG,EAAE;EACzD,CAAC;AACF,QAAO;EACL,wBAAwB,QAAQ,eAAe;AAC7C,UAAO,iBAAiB,SAAS,QAAQ,cAAc;;EAEzD,UAAU;AAER,OAAI;AACF,aAAS,WAAW;WACd;AAGR,WAAQ,WAAW;;EAEtB;;;;;;;;AASH,SAAS,iBACP,SACA,QACA,eAKuB;CACvB,MAAMC,QAAkB,EAAE;CAC1B,IAAI,OAAO;CACX,IAAIC;CACJ,IAAIC,OAA4B;CAChC,MAAM,eAAqB;AACzB,MAAI,SAAS,MAAM;GACjB,MAAM,IAAI;AACV,UAAO;AACP,MAAG;;;AAGP,CAAK,QACF,OAAO,QAAQ;EACd,GAAI,eAAe,WAAW,SAAY,EAAE,QAAQ,cAAc,QAAQ,GAAG,EAAE;EAC/E,GAAI,eAAe,cAAc,SAAY,EAAE,WAAW,cAAc,WAAW,GAAG,EAAE;EACxF,GAAI,eAAe,gBAAgB,SAC/B,EAAE,aAAa,cAAc,aAAa,GAC1C,EAAE;EACN,cAAc,UAAkB;AAC9B,OAAI,OAAO,UAAU,YAAY,MAAM,SAAS,EAAG,OAAM,KAAK,MAAM;AACpE,WAAQ;;EAEX,CAAC,CACD,WACO;AACJ,SAAO;AACP,UAAQ;KAET,QAAiB;AAChB,YAAU,eAAe,QAAQ,MAAM,IAAI,MAAM,OAAO,IAAI,CAAC;AAC7D,SAAO;AACP,UAAQ;GAEX;AACH,QAAO,EACL,CAAC,OAAO,iBAAwC;AAC9C,SAAO,EACL,MAAM,OAAwC;AAC5C,YAAS;IACP,MAAM,OAAO,MAAM,OAAO;AAC1B,QAAI,SAAS,OAAW,QAAO;KAAE,MAAM;KAAO,OAAO;KAAM;AAC3D,QAAI,YAAY,OAAW,OAAM;AACjC,QAAI,KAAM,QAAO;KAAE,MAAM;KAAM,OAAO;KAAW;AACjD,UAAM,IAAI,SAAe,YAAY;AACnC,YAAO;MACP;;KAGP;IAEJ;;AAGH,SAAS,aAAa,KAA8B;CAClD,MAAMC,QAAkB,EAAE;AAK1B,MAAK,MAAM,OAAO,IAAI,UAAU;EAC9B,MAAM,OACJ,OAAO,IAAI,YAAY,WACnB,IAAI,UACJ,IAAI,QACD,KAAK,MAAO,EAAE,SAAS,SAAS,EAAE,OAAO,EAAE,SAAS,cAAc,EAAE,OAAO,GAAI,CAC/E,KAAK,GAAG;AACjB,QAAM,KAAK,IAAI,IAAI,KAAK,IAAI,OAAO;;AAErC,QAAO,MAAM,KAAK,KAAK;;AAGzB,SAAS,SAAS,MAAsB;CACtC,MAAM,MAAM,KAAK,IAAI,KAAK,YAAY,IAAI,EAAE,KAAK,YAAY,KAAK,CAAC;AACnE,QAAO,QAAQ,KAAK,OAAO,KAAK,MAAM,MAAM,EAAE;;AAGhD,SAAS,SAAS,QAAiF;AACjG,KAAI,WAAW,QAAQ,WAAW,OAAW,QAAO;AACpD,QAAQ,OAA+B,UAAU"}

package/dist/index.d.ts

@@ -25,8 +25,7 @@ import { LlamaCppNativeCounter, LlamaCppNativeCounterOptions } from "./counter.j
  *
  * @packageDocumentation
  */
-/** Canonical version constant. Mirrors the `package.json` version. */
-declare const VERSION = "0.6.0";
+declare const VERSION: string;
 //#endregion
 export { LlamaCppNativeCounter, type LlamaCppNativeCounterOptions, type LlamaCppNodeAdapterOptions, type LlamaCppNodeRuntimeOverrides, type LlamaInstance, type LlamaModelInstance, type LlamaSessionInstance, VERSION, llamaCppNodeAdapter };
 //# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","names":[],"sources":["../src/index.ts"],"sourcesContent":[],"mappings":";;;;;;;;;;;AAuBA;;;;;;;;;;;;;;;;;cAAa,OAAA"}
+{"version":3,"file":"index.d.ts","names":[],"sources":["../src/index.ts"],"sourcesContent":[],"mappings":";;;;;;;;;;;AAyBA;;;;;;;;;;;;;;;;cAAa"}

package/dist/index.js

@@ -1,3 +1,4 @@
+import { version } from "./package.js";
 import { llamaCppNodeAdapter } from "./adapter.js";
 import { LlamaCppNativeCounter } from "./counter.js";
 
@@ -23,8 +24,8 @@ import { LlamaCppNativeCounter } from "./counter.js";
 *
 * @packageDocumentation
 */
-/** Canonical version constant. Mirrors the `package.json` version. */
-const VERSION = "0.6.0";
+/** Canonical version constant, derived from `package.json` at build time. */
+const VERSION = version;
 
 //#endregion
 export { LlamaCppNativeCounter, VERSION, llamaCppNodeAdapter };

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","names":[],"sources":["../src/index.ts"],"sourcesContent":["/**\n * @graphorin/provider-llamacpp-node - in-process GGUF execution\n * adapter for the Graphorin framework. The package wraps\n * `node-llama-cpp@^3.5` to load `.gguf` model files directly into the\n * same Node process - no daemon, no port to manage, no GPU contention\n * with other processes.\n *\n * The adapter declares `trust: 'loopback'` permanently because the\n * model lives in the same trust boundary as the host process; the\n * symmetry mirrors `@graphorin/embedder-transformersjs` (in-process\n * embedder; same trust boundary).\n *\n * The companion package is operationally simpler than the HTTP-shaped\n * adapters but does NOT survive a process restart mid-stream - the\n * model context lives in the process and is lost on exit. For HITL\n * durable mid-stream resume, one of the HTTP-shaped adapters\n * (`ollamaAdapter`, `llamaCppServerAdapter`, `openAICompatibleAdapter`)\n * is the better choice.\n *\n * @packageDocumentation\n */\n\n/** Canonical version constant. Mirrors the `package.json` version. */\nexport const VERSION = '0.6.0';\n\nexport { type LlamaCppNodeAdapterOptions, llamaCppNodeAdapter } from './adapter.js';\nexport {\n  LlamaCppNativeCounter,\n  type LlamaCppNativeCounterOptions,\n} from './counter.js';\nexport type {\n  LlamaCppNodeRuntimeOverrides,\n  LlamaInstance,\n  LlamaModelInstance,\n  LlamaSessionInstance,\n} from './runtime.js';\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;AAuBA,MAAa,UAAU"}
+{"version":3,"file":"index.js","names":["VERSION: string","pkg.version"],"sources":["../src/index.ts"],"sourcesContent":["/**\n * @graphorin/provider-llamacpp-node - in-process GGUF execution\n * adapter for the Graphorin framework. The package wraps\n * `node-llama-cpp@^3.5` to load `.gguf` model files directly into the\n * same Node process - no daemon, no port to manage, no GPU contention\n * with other processes.\n *\n * The adapter declares `trust: 'loopback'` permanently because the\n * model lives in the same trust boundary as the host process; the\n * symmetry mirrors `@graphorin/embedder-transformersjs` (in-process\n * embedder; same trust boundary).\n *\n * The companion package is operationally simpler than the HTTP-shaped\n * adapters but does NOT survive a process restart mid-stream - the\n * model context lives in the process and is lost on exit. For HITL\n * durable mid-stream resume, one of the HTTP-shaped adapters\n * (`ollamaAdapter`, `llamaCppServerAdapter`, `openAICompatibleAdapter`)\n * is the better choice.\n *\n * @packageDocumentation\n */\n\n/** Canonical version constant, derived from `package.json` at build time. */\nimport pkg from '../package.json' with { type: 'json' };\n\nexport const VERSION: string = pkg.version;\n\nexport { type LlamaCppNodeAdapterOptions, llamaCppNodeAdapter } from './adapter.js';\nexport {\n  LlamaCppNativeCounter,\n  type LlamaCppNativeCounterOptions,\n} from './counter.js';\nexport type {\n  LlamaCppNodeRuntimeOverrides,\n  LlamaInstance,\n  LlamaModelInstance,\n  LlamaSessionInstance,\n} from './runtime.js';\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;AAyBA,MAAaA,UAAkBC"}

package/dist/package.js

@@ -0,0 +1,6 @@
+//#region package.json
+var version = "0.7.0";
+
+//#endregion
+export { version };
+//# sourceMappingURL=package.js.map

package/dist/package.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"package.js","names":[],"sources":["../package.json"],"sourcesContent":["{\n  \"name\": \"@graphorin/provider-llamacpp-node\",\n  \"version\": \"0.7.0\",\n  \"description\": \"In-process GGUF execution adapter for the Graphorin framework. Wraps node-llama-cpp@^3.5 to load .gguf model files directly into the same Node process - no daemon, no port to manage, no GPU contention with other processes. Declares trust: 'loopback' permanently because the model lives in the same trust boundary as the host process. Ships LlamaCppNativeCounter that wraps model.tokenize() from the loaded GGUF for byte-exact token counts.\",\n  \"license\": \"MIT\",\n  \"author\": \"Oleksiy Stepurenko\",\n  \"homepage\": \"https://github.com/o-stepper/graphorin/tree/main/packages/provider-llamacpp-node\",\n  \"repository\": {\n    \"type\": \"git\",\n    \"url\": \"git+https://github.com/o-stepper/graphorin.git\",\n    \"directory\": \"packages/provider-llamacpp-node\"\n  },\n  \"bugs\": {\n    \"url\": \"https://github.com/o-stepper/graphorin/issues\"\n  },\n  \"keywords\": [\n    \"graphorin\",\n    \"ai\",\n    \"agents\",\n    \"framework\",\n    \"provider\",\n    \"llm\",\n    \"llama-cpp\",\n    \"node-llama-cpp\",\n    \"gguf\",\n    \"in-process\",\n    \"local-first\",\n    \"loopback\"\n  ],\n  \"type\": \"module\",\n  \"sideEffects\": false,\n  \"engines\": {\n    \"node\": \">=22.12.0\"\n  },\n  \"main\": \"./dist/index.js\",\n  \"module\": \"./dist/index.js\",\n  \"types\": \"./dist/index.d.ts\",\n  \"exports\": {\n    \".\": {\n      \"types\": \"./dist/index.d.ts\",\n      \"default\": \"./dist/index.js\"\n    },\n    \"./package.json\": \"./package.json\"\n  },\n  \"files\": [\n    \"dist\",\n    \"src\",\n    \"README.md\",\n    \"CHANGELOG.md\",\n    \"LICENSE\"\n  ],\n  \"scripts\": {\n    \"build\": \"tsdown\",\n    \"typecheck\": \"tsc --noEmit && tsc -p tsconfig.tests.json\",\n    \"test\": \"vitest run\",\n    \"lint\": \"biome check .\",\n    \"clean\": \"rimraf dist .turbo *.tsbuildinfo\"\n  },\n  \"dependencies\": {\n    \"@graphorin/core\": \"workspace:*\",\n    \"@graphorin/provider\": \"workspace:*\"\n  },\n  \"peerDependencies\": {\n    \"node-llama-cpp\": \"^3.5.0\"\n  },\n  \"peerDependenciesMeta\": {\n    \"node-llama-cpp\": {\n      \"optional\": false\n    }\n  },\n  \"publishConfig\": {\n    \"access\": \"public\",\n    \"provenance\": true\n  },\n  \"devDependencies\": {}\n}\n"],"mappings":";cAEa"}

package/dist/runtime.d.ts

@@ -35,6 +35,22 @@ interface LlamaModelInstance {
   }>;
   dispose?(): Promise<void>;
 }
+/**
+ * Structural mirror of node-llama-cpp v3's `ChatHistoryItem` (W-096).
+ * A `'model'` turn carries its text as `response: string[]`.
+ *
+ * @internal
+ */
+type LlamaChatHistoryItem = {
+  readonly type: 'system';
+  readonly text: string;
+} | {
+  readonly type: 'user';
+  readonly text: string;
+} | {
+  readonly type: 'model';
+  readonly response: ReadonlyArray<string>;
+};
 /**
  * Loaded chat session capable of streaming responses.
  *
@@ -46,6 +62,15 @@ interface LlamaSessionInstance {
     readonly maxTokens?: number;
     readonly temperature?: number;
   }): AsyncIterable<string>;
+  /**
+   * W-096: replace the session's chat history (node-llama-cpp v3
+   * `setChatHistory`). When present, the adapter feeds multi-turn
+   * transcripts as REAL chat history + prompts only the last user turn
+   * - instead of serialising the whole conversation into one
+   * pseudo-prompt string. Optional: fixtures / custom factories without
+   * it keep the legacy render-prompt path.
+   */
+  setChatHistory?(history: ReadonlyArray<LlamaChatHistoryItem>): void;
   /**
    * Release the per-request context / sequence backing this session.
    * node-llama-cpp contexts hold KV-cache memory (hundreds of MB at
@@ -90,6 +115,8 @@ interface LlamaChatSessionPeer {
     readonly temperature?: number;
     readonly onTextChunk?: (chunk: string) => void;
   }): Promise<string>;
+  /** W-096: node-llama-cpp v3 chat-history setter (optional slice). */
+  setChatHistory?(history: ReadonlyArray<LlamaChatHistoryItem>): void;
 }
 /** @internal */
 type LlamaChatSessionCtor = new (args: {

package/dist/runtime.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"runtime.d.ts","names":[],"sources":["../src/runtime.ts"],"sourcesContent":[],"mappings":";;AAaA;AASA;;;;;;AAeA;AAuBA;;;AASW,UAxDM,aAAA,CAwDN;EAEI,SAAA,CAAA,IAAA,EAAA;IAAR,SAAA,EAAA,MAAA;IAMuB,SAAA,CAAA,EAAA,MAAA,GAAA,MAAA;EAAoB,CAAA,CAAA,EA/DqB,OA+DrB,CA/D6B,kBA+D7B,CAAA;AAWlD;AAaA;;;;;UA/EiB,kBAAA;;8CAE6B,cAAc;;;MACV;;;;;;cAIpC;;;;;;;UAQG,oBAAA;;sBAIO;;;MAInB;;;;;;;;;;;;;;UAeY,4BAAA;;4BAEW,QAAQ;;;;;;mCAOzB,wCAEJ,QAAQ;;;;;;8BAMe;;;;;;;;;;UAWb,oBAAA;;sBAIO;;;;MAKnB;;;KAIO,oBAAA;;;MAGN"}
+{"version":3,"file":"runtime.d.ts","names":[],"sources":["../src/runtime.ts"],"sourcesContent":[],"mappings":";;AAaA;AASA;;;;;;AAgBA;AAUA;;;AAiByC,UApDxB,aAAA,CAoDwB;EAAd,SAAA,CAAA,IAAA,EAAA;IAAa,SAAA,EAAA,MAAA;IAevB,SAAA,CAAA,EAAA,MAAA,GAAA,MAA4B;EAET,CAAA,CAAA,EApEmC,OAoEnC,CApE2C,kBAoE3C,CAAA;;;;;;;AA0BnB,UAtFA,kBAAA,CAsFoB;EAIb,SAAA,gBAAA,CAAA,EAAA,MAAA;EAKnB,QAAA,CAAA,IAAA,EAAA,MAAA,CAAA,EAAA,SAAA,MAAA,EAAA,GA7FyC,WA6FzC,GA7FuD,UA6FvD;EAEoC,aAAA,CAAA,IAAD,CAAC,EAAA;IAAd,WAAA,CAAA,EAAA,MAAA;EAAa,CAAA,CAAA,EA9FU,OA8FV,CAAA;IAI5B,WAAA,EAAA,EAAA;;;;;cA9FE;;;;;;;;KASF,oBAAA;;;;;;;;qBAGqC;;;;;;;UAOhC,oBAAA;;sBAIO;;;MAInB;;;;;;;;;2BASsB,cAAc;;;;;;;;;;;;;;UAexB,4BAAA;;4BAEW,QAAQ;;;;;;mCAOzB,wCAEJ,QAAQ;;;;;;8BAMe;;;;;;;;;;UAWb,oBAAA;;sBAIO;;;;MAKnB;;2BAEsB,cAAc;;;KAI7B,oBAAA;;;MAGN"}

package/dist/runtime.js.map

@@ -1 +1 @@
-{"version":3,"file":"runtime.js","names":["cachedLlama: LlamaInstance | null","cachedChatSessionCtor: LlamaChatSessionCtor | null","mod: { getLlama?: unknown }","mod: { LlamaChatSession?: unknown }"],"sources":["../src/runtime.ts"],"sourcesContent":["/**\n * Loose structural shapes covering the slice of `node-llama-cpp` we\n * use. Re-declared here to avoid pulling the heavy native peer at\n * type-check time.\n *\n * @internal\n */\n\n/**\n * `Llama` engine instance (returned by `getLlama()`).\n *\n * @internal\n */\nexport interface LlamaInstance {\n  loadModel(args: { modelPath: string; gpuLayers?: number | 'auto' }): Promise<LlamaModelInstance>;\n}\n\n/**\n * Loaded GGUF model.\n *\n * @internal\n */\nexport interface LlamaModelInstance {\n  readonly trainContextSize?: number;\n  tokenize(text: string): readonly number[] | Uint32Array | Uint8Array;\n  createContext(args?: { contextSize?: number }): Promise<{\n    getSequence(): { dispose?: () => void };\n    dispose?: () => void;\n  }>;\n  dispose?(): Promise<void>;\n}\n\n/**\n * Loaded chat session capable of streaming responses.\n *\n * @internal\n */\nexport interface LlamaSessionInstance {\n  promptStreamingResponse(\n    prompt: string,\n    options?: {\n      readonly signal?: AbortSignal;\n      readonly maxTokens?: number;\n      readonly temperature?: number;\n    },\n  ): AsyncIterable<string>;\n  /**\n   * Release the per-request context / sequence backing this session.\n   * node-llama-cpp contexts hold KV-cache memory (hundreds of MB at\n   * large context sizes); the adapter calls this in a `finally` after\n   * every stream so long-running agents do not leak until OOM.\n   */\n  dispose?(): void;\n}\n\n/**\n * Test-only shape for injecting fixture-driven runtime behaviour.\n *\n * @stable\n */\nexport interface LlamaCppNodeRuntimeOverrides {\n  /** Returns a `LlamaInstance` (the result of `getLlama()`). */\n  readonly getLlama?: () => Promise<LlamaInstance>;\n  /**\n   * Build a streaming chat session against an already-loaded model\n   * instance. Used by the adapter to wire `model.tokenize` and\n   * `session.promptStreamingResponse` to the per-test fixture.\n   */\n  readonly createSession?: (\n    model: LlamaModelInstance,\n    system?: string,\n  ) => Promise<LlamaSessionInstance>;\n  /**\n   * Override the `LlamaChatSession` constructor used by the REAL\n   * default session factory (PS-3). Tests stub it; production loads it\n   * from the `node-llama-cpp` peer.\n   */\n  readonly LlamaChatSession?: LlamaChatSessionCtor;\n}\n\n/**\n * Structural slice of the peer's `LlamaChatSession` class used by the\n * default session factory (PS-3): `prompt(text, { onTextChunk })`\n * resolves with the full response while streaming chunks through the\n * callback.\n *\n * @internal\n */\nexport interface LlamaChatSessionPeer {\n  prompt(\n    text: string,\n    options?: {\n      readonly signal?: AbortSignal;\n      readonly maxTokens?: number;\n      readonly temperature?: number;\n      readonly onTextChunk?: (chunk: string) => void;\n    },\n  ): Promise<string>;\n}\n\n/** @internal */\nexport type LlamaChatSessionCtor = new (args: {\n  readonly contextSequence: unknown;\n  readonly systemPrompt?: string;\n}) => LlamaChatSessionPeer;\n\nlet cachedLlama: LlamaInstance | null = null;\nlet cachedChatSessionCtor: LlamaChatSessionCtor | null = null;\n\n/**\n * Lazily load the `node-llama-cpp` peer and return the `Llama` engine\n * instance. Cached per process.\n *\n * @internal\n */\nexport async function loadLlamaModule(\n  overrides: LlamaCppNodeRuntimeOverrides | undefined,\n): Promise<LlamaInstance> {\n  if (overrides?.getLlama !== undefined) return overrides.getLlama();\n  if (cachedLlama !== null) return cachedLlama;\n  let mod: { getLlama?: unknown };\n  try {\n    mod = (await import('node-llama-cpp')) as { getLlama?: unknown };\n  } catch (cause) {\n    throw new Error(\n      \"[graphorin/provider-llamacpp-node] missing peer dependency 'node-llama-cpp'. \" +\n        'Install it with `pnpm add node-llama-cpp`.',\n      { cause },\n    );\n  }\n  if (typeof mod.getLlama !== 'function') {\n    throw new Error(\n      '[graphorin/provider-llamacpp-node] installed node-llama-cpp does not expose getLlama().',\n    );\n  }\n  const instance = (await (mod.getLlama as () => Promise<LlamaInstance>)()) as LlamaInstance;\n  cachedLlama = instance;\n  return instance;\n}\n\n/**\n * Lazily resolve the peer's `LlamaChatSession` constructor for the\n * real default session factory (PS-3). Cached per process; the\n * override wins for tests.\n *\n * @internal\n */\nexport async function loadLlamaChatSessionCtor(\n  overrides: LlamaCppNodeRuntimeOverrides | undefined,\n): Promise<LlamaChatSessionCtor> {\n  if (overrides?.LlamaChatSession !== undefined) return overrides.LlamaChatSession;\n  if (cachedChatSessionCtor !== null) return cachedChatSessionCtor;\n  let mod: { LlamaChatSession?: unknown };\n  try {\n    mod = (await import('node-llama-cpp')) as { LlamaChatSession?: unknown };\n  } catch (cause) {\n    throw new Error(\n      \"[graphorin/provider-llamacpp-node] missing peer dependency 'node-llama-cpp'. \" +\n        'Install it with `pnpm add node-llama-cpp`.',\n      { cause },\n    );\n  }\n  if (typeof mod.LlamaChatSession !== 'function') {\n    throw new Error(\n      '[graphorin/provider-llamacpp-node] installed node-llama-cpp does not expose LlamaChatSession.',\n    );\n  }\n  cachedChatSessionCtor = mod.LlamaChatSession as LlamaChatSessionCtor;\n  return cachedChatSessionCtor;\n}\n\n/**\n * Test-only hook that resets the cached `getLlama()` result.\n *\n * @internal\n */\nexport function __resetLlamaCache(): void {\n  cachedLlama = null;\n  cachedChatSessionCtor = null;\n}\n"],"mappings":";AA0GA,IAAIA,cAAoC;AACxC,IAAIC,wBAAqD;;;;;;;AAQzD,eAAsB,gBACpB,WACwB;AACxB,KAAI,WAAW,aAAa,OAAW,QAAO,UAAU,UAAU;AAClE,KAAI,gBAAgB,KAAM,QAAO;CACjC,IAAIC;AACJ,KAAI;AACF,QAAO,MAAM,OAAO;UACb,OAAO;AACd,QAAM,IAAI,MACR,2HAEA,EAAE,OAAO,CACV;;AAEH,KAAI,OAAO,IAAI,aAAa,WAC1B,OAAM,IAAI,MACR,0FACD;CAEH,MAAM,WAAY,MAAO,IAAI,UAA2C;AACxE,eAAc;AACd,QAAO;;;;;;;;;AAUT,eAAsB,yBACpB,WAC+B;AAC/B,KAAI,WAAW,qBAAqB,OAAW,QAAO,UAAU;AAChE,KAAI,0BAA0B,KAAM,QAAO;CAC3C,IAAIC;AACJ,KAAI;AACF,QAAO,MAAM,OAAO;UACb,OAAO;AACd,QAAM,IAAI,MACR,2HAEA,EAAE,OAAO,CACV;;AAEH,KAAI,OAAO,IAAI,qBAAqB,WAClC,OAAM,IAAI,MACR,gGACD;AAEH,yBAAwB,IAAI;AAC5B,QAAO"}
+{"version":3,"file":"runtime.js","names":["cachedLlama: LlamaInstance | null","cachedChatSessionCtor: LlamaChatSessionCtor | null","mod: { getLlama?: unknown }","mod: { LlamaChatSession?: unknown }"],"sources":["../src/runtime.ts"],"sourcesContent":["/**\n * Loose structural shapes covering the slice of `node-llama-cpp` we\n * use. Re-declared here to avoid pulling the heavy native peer at\n * type-check time.\n *\n * @internal\n */\n\n/**\n * `Llama` engine instance (returned by `getLlama()`).\n *\n * @internal\n */\nexport interface LlamaInstance {\n  loadModel(args: { modelPath: string; gpuLayers?: number | 'auto' }): Promise<LlamaModelInstance>;\n}\n\n/**\n * Loaded GGUF model.\n *\n * @internal\n */\nexport interface LlamaModelInstance {\n  readonly trainContextSize?: number;\n  tokenize(text: string): readonly number[] | Uint32Array | Uint8Array;\n  createContext(args?: { contextSize?: number }): Promise<{\n    getSequence(): { dispose?: () => void };\n    dispose?: () => void;\n  }>;\n  dispose?(): Promise<void>;\n}\n\n/**\n * Structural mirror of node-llama-cpp v3's `ChatHistoryItem` (W-096).\n * A `'model'` turn carries its text as `response: string[]`.\n *\n * @internal\n */\nexport type LlamaChatHistoryItem =\n  | { readonly type: 'system'; readonly text: string }\n  | { readonly type: 'user'; readonly text: string }\n  | { readonly type: 'model'; readonly response: ReadonlyArray<string> };\n\n/**\n * Loaded chat session capable of streaming responses.\n *\n * @internal\n */\nexport interface LlamaSessionInstance {\n  promptStreamingResponse(\n    prompt: string,\n    options?: {\n      readonly signal?: AbortSignal;\n      readonly maxTokens?: number;\n      readonly temperature?: number;\n    },\n  ): AsyncIterable<string>;\n  /**\n   * W-096: replace the session's chat history (node-llama-cpp v3\n   * `setChatHistory`). When present, the adapter feeds multi-turn\n   * transcripts as REAL chat history + prompts only the last user turn\n   * - instead of serialising the whole conversation into one\n   * pseudo-prompt string. Optional: fixtures / custom factories without\n   * it keep the legacy render-prompt path.\n   */\n  setChatHistory?(history: ReadonlyArray<LlamaChatHistoryItem>): void;\n  /**\n   * Release the per-request context / sequence backing this session.\n   * node-llama-cpp contexts hold KV-cache memory (hundreds of MB at\n   * large context sizes); the adapter calls this in a `finally` after\n   * every stream so long-running agents do not leak until OOM.\n   */\n  dispose?(): void;\n}\n\n/**\n * Test-only shape for injecting fixture-driven runtime behaviour.\n *\n * @stable\n */\nexport interface LlamaCppNodeRuntimeOverrides {\n  /** Returns a `LlamaInstance` (the result of `getLlama()`). */\n  readonly getLlama?: () => Promise<LlamaInstance>;\n  /**\n   * Build a streaming chat session against an already-loaded model\n   * instance. Used by the adapter to wire `model.tokenize` and\n   * `session.promptStreamingResponse` to the per-test fixture.\n   */\n  readonly createSession?: (\n    model: LlamaModelInstance,\n    system?: string,\n  ) => Promise<LlamaSessionInstance>;\n  /**\n   * Override the `LlamaChatSession` constructor used by the REAL\n   * default session factory (PS-3). Tests stub it; production loads it\n   * from the `node-llama-cpp` peer.\n   */\n  readonly LlamaChatSession?: LlamaChatSessionCtor;\n}\n\n/**\n * Structural slice of the peer's `LlamaChatSession` class used by the\n * default session factory (PS-3): `prompt(text, { onTextChunk })`\n * resolves with the full response while streaming chunks through the\n * callback.\n *\n * @internal\n */\nexport interface LlamaChatSessionPeer {\n  prompt(\n    text: string,\n    options?: {\n      readonly signal?: AbortSignal;\n      readonly maxTokens?: number;\n      readonly temperature?: number;\n      readonly onTextChunk?: (chunk: string) => void;\n    },\n  ): Promise<string>;\n  /** W-096: node-llama-cpp v3 chat-history setter (optional slice). */\n  setChatHistory?(history: ReadonlyArray<LlamaChatHistoryItem>): void;\n}\n\n/** @internal */\nexport type LlamaChatSessionCtor = new (args: {\n  readonly contextSequence: unknown;\n  readonly systemPrompt?: string;\n}) => LlamaChatSessionPeer;\n\nlet cachedLlama: LlamaInstance | null = null;\nlet cachedChatSessionCtor: LlamaChatSessionCtor | null = null;\n\n/**\n * Lazily load the `node-llama-cpp` peer and return the `Llama` engine\n * instance. Cached per process.\n *\n * @internal\n */\nexport async function loadLlamaModule(\n  overrides: LlamaCppNodeRuntimeOverrides | undefined,\n): Promise<LlamaInstance> {\n  if (overrides?.getLlama !== undefined) return overrides.getLlama();\n  if (cachedLlama !== null) return cachedLlama;\n  let mod: { getLlama?: unknown };\n  try {\n    mod = (await import('node-llama-cpp')) as { getLlama?: unknown };\n  } catch (cause) {\n    throw new Error(\n      \"[graphorin/provider-llamacpp-node] missing peer dependency 'node-llama-cpp'. \" +\n        'Install it with `pnpm add node-llama-cpp`.',\n      { cause },\n    );\n  }\n  if (typeof mod.getLlama !== 'function') {\n    throw new Error(\n      '[graphorin/provider-llamacpp-node] installed node-llama-cpp does not expose getLlama().',\n    );\n  }\n  const instance = (await (mod.getLlama as () => Promise<LlamaInstance>)()) as LlamaInstance;\n  cachedLlama = instance;\n  return instance;\n}\n\n/**\n * Lazily resolve the peer's `LlamaChatSession` constructor for the\n * real default session factory (PS-3). Cached per process; the\n * override wins for tests.\n *\n * @internal\n */\nexport async function loadLlamaChatSessionCtor(\n  overrides: LlamaCppNodeRuntimeOverrides | undefined,\n): Promise<LlamaChatSessionCtor> {\n  if (overrides?.LlamaChatSession !== undefined) return overrides.LlamaChatSession;\n  if (cachedChatSessionCtor !== null) return cachedChatSessionCtor;\n  let mod: { LlamaChatSession?: unknown };\n  try {\n    mod = (await import('node-llama-cpp')) as { LlamaChatSession?: unknown };\n  } catch (cause) {\n    throw new Error(\n      \"[graphorin/provider-llamacpp-node] missing peer dependency 'node-llama-cpp'. \" +\n        'Install it with `pnpm add node-llama-cpp`.',\n      { cause },\n    );\n  }\n  if (typeof mod.LlamaChatSession !== 'function') {\n    throw new Error(\n      '[graphorin/provider-llamacpp-node] installed node-llama-cpp does not expose LlamaChatSession.',\n    );\n  }\n  cachedChatSessionCtor = mod.LlamaChatSession as LlamaChatSessionCtor;\n  return cachedChatSessionCtor;\n}\n\n/**\n * Test-only hook that resets the cached `getLlama()` result.\n *\n * @internal\n */\nexport function __resetLlamaCache(): void {\n  cachedLlama = null;\n  cachedChatSessionCtor = null;\n}\n"],"mappings":";AAgIA,IAAIA,cAAoC;AACxC,IAAIC,wBAAqD;;;;;;;AAQzD,eAAsB,gBACpB,WACwB;AACxB,KAAI,WAAW,aAAa,OAAW,QAAO,UAAU,UAAU;AAClE,KAAI,gBAAgB,KAAM,QAAO;CACjC,IAAIC;AACJ,KAAI;AACF,QAAO,MAAM,OAAO;UACb,OAAO;AACd,QAAM,IAAI,MACR,2HAEA,EAAE,OAAO,CACV;;AAEH,KAAI,OAAO,IAAI,aAAa,WAC1B,OAAM,IAAI,MACR,0FACD;CAEH,MAAM,WAAY,MAAO,IAAI,UAA2C;AACxE,eAAc;AACd,QAAO;;;;;;;;;AAUT,eAAsB,yBACpB,WAC+B;AAC/B,KAAI,WAAW,qBAAqB,OAAW,QAAO,UAAU;AAChE,KAAI,0BAA0B,KAAM,QAAO;CAC3C,IAAIC;AACJ,KAAI;AACF,QAAO,MAAM,OAAO;UACb,OAAO;AACd,QAAM,IAAI,MACR,2HAEA,EAAE,OAAO,CACV;;AAEH,KAAI,OAAO,IAAI,qBAAqB,WAClC,OAAM,IAAI,MACR,gGACD;AAEH,yBAAwB,IAAI;AAC5B,QAAO"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@graphorin/provider-llamacpp-node",
-  "version": "0.6.0",
+  "version": "0.7.0",
   "description": "In-process GGUF execution adapter for the Graphorin framework. Wraps node-llama-cpp@^3.5 to load .gguf model files directly into the same Node process - no daemon, no port to manage, no GPU contention with other processes. Declares trust: 'loopback' permanently because the model lives in the same trust boundary as the host process. Ships LlamaCppNativeCounter that wraps model.tokenize() from the loaded GGUF for byte-exact token counts.",
   "license": "MIT",
   "author": "Oleksiy Stepurenko",
@@ -28,8 +28,9 @@
     "loopback"
   ],
   "type": "module",
+  "sideEffects": false,
   "engines": {
-    "node": ">=22.0.0"
+    "node": ">=22.12.0"
   },
   "main": "./dist/index.js",
   "module": "./dist/index.js",
@@ -37,19 +38,20 @@
   "exports": {
     ".": {
       "types": "./dist/index.d.ts",
-      "import": "./dist/index.js"
+      "default": "./dist/index.js"
     },
     "./package.json": "./package.json"
   },
   "files": [
     "dist",
+    "src",
     "README.md",
     "CHANGELOG.md",
     "LICENSE"
   ],
   "dependencies": {
-    "@graphorin/core": "0.6.0",
-    "@graphorin/provider": "0.6.0"
+    "@graphorin/core": "0.7.0",
+    "@graphorin/provider": "0.7.0"
   },
   "peerDependencies": {
     "node-llama-cpp": "^3.5.0"