@oh-my-pi/pi-coding-agent 17.1.7 → 17.2.0

package/src/session/session-entries.ts

@@ -154,6 +154,7 @@ export interface TitleChangeEntry extends SessionEntryBase {
 declare module "@oh-my-pi/pi-agent-core/compaction/entries" {
 	interface CustomCompactionSessionEntries {
 		titleChange: TitleChangeEntry;
+		credentialPin: CredentialPinEntry;
 	}
 }
 
@@ -164,6 +165,24 @@ export interface TtsrInjectionEntry extends SessionEntryBase {
 	injectedRules: string[];
 }
 
+/**
+ * Records which OAuth account served this session's requests for a provider.
+ *
+ * Provider prompt caches (Anthropic in particular) are account-scoped, and the
+ * auth store's session-sticky routing is process-local under a remote auth
+ * broker, so resume must re-pin the same account to reuse the warm cache
+ * prefix. Stores a sha-256 of the account + billing-scope tuple instead of
+ * the raw email/uuid/org; note an unsalted digest of a guessable email is
+ * still linkable, so exported sessions are pseudonymous, not anonymous.
+ */
+export interface CredentialPinEntry extends SessionEntryBase {
+	type: "credential_pin";
+	/** Provider id the pin applies to (e.g. "anthropic"). */
+	provider: string;
+	/** `credentialPinHash()` of the serving account's identity + scope tuple. */
+	hash: string;
+}
+
 /** Session init entry - captures initial context for subagent sessions (debugging/replay). */
 export interface SessionInitEntry extends SessionEntryBase {
 	type: "session_init";
@@ -230,7 +249,8 @@ export type SessionEntry =
 	| TitleChangeEntry
 	| TtsrInjectionEntry
 	| SessionInitEntry
-	| ModeChangeEntry;
+	| ModeChangeEntry
+	| CredentialPinEntry;
 
 /** Raw logical file entry after loaders strip any fixed-width title slot. */
 export type FileEntry = SessionHeader | SessionEntry;

package/src/session/session-handoff.ts

@@ -64,7 +64,10 @@ export interface SessionHandoffHost {
 	clearPendingNextTurnMessages(): void;
 	resetTodoCycle(): void;
 	buildDisplaySessionContext(): SessionContext;
-	resetAdvisorRuntimes(): void;
+	resetAdvisorSessionState(): void;
+	drainAndDetachAdvisorRecorders(): Promise<void>;
+	reattachAdvisorRecorderFeeds(): void;
+	clearAdvisorCost(): void;
 	syncTodoPhasesFromBranch(): void;
 }
 
@@ -124,6 +127,8 @@ export class SessionHandoff {
 			}
 		}
 
+		let advisorRecordersDetached = false;
+		let sessionTransitioned = false;
 		try {
 			if (handoffSignal.aborted) {
 				throw new Error("Handoff cancelled");
@@ -224,14 +229,21 @@ export class SessionHandoff {
 			}
 			await this.#host.flushPendingBash();
 			await this.#host.sessionManager.flush();
+			advisorRecordersDetached = true;
+			// Stop and settle in-flight advisors while the old-session feeds can still
+			// observe message_end, then mute before opening the replacement session.
+			await this.#host.drainAndDetachAdvisorRecorders();
 			const bashTransition = this.#host.beginBashSessionTransition();
 			this.#host.cancelOwnAsyncJobs();
-			let sessionTransitioned = false;
 			try {
 				await this.#host.sessionManager.newSession(
 					previousSessionFile ? { parentSession: previousSessionFile } : undefined,
 				);
 				this.#host.markBashSessionTransition(bashTransition);
+				// The handoff opens a fresh conversation, so the spend of the one it
+				// summarizes stays with it. Clearing here, at the commit point, keeps the
+				// status line honest even if a later step throws.
+				this.#host.clearAdvisorCost();
 				sessionTransitioned = true;
 			} finally {
 				this.#host.finishBashSessionTransition(bashTransition, sessionTransitioned);
@@ -284,7 +296,8 @@ export class SessionHandoff {
 			// Rebuild agent messages from session
 			const sessionContext = this.#host.buildDisplaySessionContext();
 			this.#host.agent.replaceMessages(sessionContext.messages);
-			this.#host.resetAdvisorRuntimes();
+			this.#host.resetAdvisorSessionState();
+			advisorRecordersDetached = false;
 			this.#host.syncTodoPhasesFromBranch();
 			if (this.#host.extensionRunner) {
 				await this.#host.extensionRunner.emit({
@@ -301,6 +314,10 @@ export class SessionHandoff {
 			}
 			throw error;
 		} finally {
+			if (advisorRecordersDetached) {
+				if (sessionTransitioned) this.#host.resetAdvisorSessionState();
+				else this.#host.reattachAdvisorRecorderFeeds();
+			}
 			sourceSignal?.removeEventListener("abort", onSourceAbort);
 			this.#handoffAbortController = undefined;
 		}

package/src/session/session-history-format.ts

@@ -218,7 +218,10 @@ function toolCallLine(
 	return base;
 }
 
-/** One line for a user-initiated `!`/`$` execution. */
+/** One line for a user-initiated `!`/`$` execution. Always attributed to the
+ *  user: these roles never carry agent-run commands (the model's bash goes
+ *  through `toolCall`), so the `user-` prefix makes provenance explicit for the
+ *  advisor and history readers regardless of render mode. */
 function executionLine(
 	kind: "bash" | "python",
 	source: string,
@@ -231,7 +234,7 @@ function executionLine(
 			: "ok";
 	const lines = lineCount(msg.output);
 	const sourcePreview = formatExecutionSourcePreview(source);
-	return `→ ${kind}! ${sourcePreview} ⇒ ${status} · ${lines} ${lines === 1 ? "line" : "lines"}`;
+	return `→ user-${kind}! ${sourcePreview} ⇒ ${status} · ${lines} ${lines === 1 ? "line" : "lines"}`;
 }
 
 /**
@@ -311,6 +314,18 @@ export function formatSessionHistoryMarkdown(messages: unknown[], opts?: History
 	// every call repeats `**agent**:`). Cleared whenever a
 	// non-role-labeled line is emitted so the next turn re-labels.
 	let lastWatchedLabel: string | undefined;
+	// Emit a watched-mode role label, collapsing consecutive same-role turns
+	// under one label (matching the user/assistant paths). Used for the
+	// user-attributed `!`/`$` execution lines so the advisor never reads them
+	// as agent actions.
+	const pushWatchedRole = (label: string, body: string): void => {
+		if (lastWatchedLabel === label) {
+			lines.push(body, "");
+		} else {
+			lines.push(label, body, "");
+			lastWatchedLabel = label;
+		}
+	};
 
 	for (const msg of typed) {
 		switch (msg.role) {
@@ -372,15 +387,25 @@ export function formatSessionHistoryMarkdown(messages: unknown[], opts?: History
 			case "bashExecution": {
 				const bashMsg = msg as BashExecutionMessage;
 				if (bashMsg.excludeFromContext) break;
-				lines.push(executionLine("bash", bashMsg.command, bashMsg), "");
-				lastWatchedLabel = undefined;
+				const bashLine = executionLine("bash", bashMsg.command, bashMsg);
+				if (opts?.watchedRoles) {
+					pushWatchedRole("**user**:", bashLine);
+				} else {
+					lines.push(bashLine, "");
+					lastWatchedLabel = undefined;
+				}
 				break;
 			}
 			case "pythonExecution": {
 				const pythonMsg = msg as PythonExecutionMessage;
 				if (pythonMsg.excludeFromContext) break;
-				lines.push(executionLine("python", pythonMsg.code, pythonMsg), "");
-				lastWatchedLabel = undefined;
+				const pythonLine = executionLine("python", pythonMsg.code, pythonMsg);
+				if (opts?.watchedRoles) {
+					pushWatchedRole("**user**:", pythonLine);
+				} else {
+					lines.push(pythonLine, "");
+					lastWatchedLabel = undefined;
+				}
 				break;
 			}
 			case "custom":

package/src/session/session-loader.ts

@@ -56,14 +56,14 @@ export function parseSessionContent(content: string): {
 	return { entries: foldTitleSlot(entries, slot), titleSlot: slot };
 }
 
-/** Exported for testing — the ≥8MiB streaming path (works on any file size). */
-export async function loadEntriesFromFileStream(filePath: string): Promise<{
-	entries: FileEntry[];
-	titleSlot: SessionTitleUpdate | undefined;
-}> {
-	const entries: FileEntry[] = [];
+/** Parse session JSONL and visit each entry without retaining prior entries. */
+export async function visitEntriesFromFileStream(
+	filePath: string,
+	visit: (entry: FileEntry) => void,
+): Promise<SessionTitleUpdate | undefined> {
 	let titleSlot: SessionTitleUpdate | undefined;
 	let sawFirstLine = false;
+	let visitorThrew = false;
 	// Byte buffer (NOT a decoded string): multibyte UTF-8 sequences that straddle
 	// a stream-chunk boundary stay intact, and Bun.JSONL.parseChunk accepts typed
 	// arrays directly. Only the unconsumed remainder is held (≤ one record + a
@@ -75,8 +75,13 @@ export async function loadEntriesFromFileStream(filePath: string): Promise<{
 	const drain = () => {
 		while (buffer.length > 0) {
 			const { values, error, read, done } = Bun.JSONL.parseChunk(buffer);
-			if (values.length > 0) {
-				for (const value of values) entries.push(value as FileEntry);
+			for (const value of values) {
+				try {
+					visit(value as FileEntry);
+				} catch (err) {
+					visitorThrew = true;
+					throw err;
+				}
 			}
 			if (error) {
 				// Malformed record: skip past the next newline and continue.
@@ -125,10 +130,21 @@ export async function loadEntriesFromFileStream(filePath: string): Promise<{
 		}
 		drain();
 	} catch (err) {
-		if (isEnoent(err)) return { entries: [], titleSlot: undefined };
+		if (visitorThrew) throw err;
+		if (isEnoent(err)) return undefined;
 		throw err;
 	}
 
+	return titleSlot;
+}
+
+/** Exported for testing — the ≥8MiB streaming path (works on any file size). */
+export async function loadEntriesFromFileStream(filePath: string): Promise<{
+	entries: FileEntry[];
+	titleSlot: SessionTitleUpdate | undefined;
+}> {
+	const entries: FileEntry[] = [];
+	const titleSlot = await visitEntriesFromFileStream(filePath, entry => entries.push(entry));
 	return { entries: foldTitleSlot(entries, titleSlot), titleSlot };
 }
 

package/src/session/session-maintenance.ts

@@ -26,6 +26,7 @@ import {
 	DEFAULT_SHAKE_CONFIG,
 	effectiveReserveTokens,
 	estimateTokens,
+	NativeCompactionError,
 	prepareCompaction,
 	resolveBudgetReserveTokens,
 	resolveThresholdTokens,
@@ -34,6 +35,7 @@ import {
 	type SummaryOptions,
 	shouldCompact,
 	shouldUseOpenAiRemoteCompaction,
+	shouldUseProviderNativeCompaction,
 } from "@oh-my-pi/pi-agent-core/compaction";
 import {
 	DEFAULT_PRUNE_CONFIG,
@@ -1373,7 +1375,11 @@ export class SessionMaintenance {
 		}
 	}
 
-	async resolveContextPromotionTarget(currentModel: Model, contextWindow: number): Promise<Model | undefined> {
+	async resolveContextPromotionTarget(
+		currentModel: Model,
+		contextWindow: number,
+		signal?: AbortSignal,
+	): Promise<Model | undefined> {
 		const availableModels = this.#host.modelRegistry.getAvailable();
 		if (availableModels.length === 0) return undefined;
 
@@ -1381,7 +1387,7 @@ export class SessionMaintenance {
 		if (!candidate) return undefined;
 		if (modelsAreEqual(candidate, currentModel)) return undefined;
 		if (candidate.contextWindow == null || candidate.contextWindow <= contextWindow) return undefined;
-		const apiKey = await this.#host.modelRegistry.getApiKey(candidate, this.#host.sessionId());
+		const apiKey = await this.#host.modelRegistry.getApiKey(candidate, this.#host.sessionId(), { signal });
 		if (!apiKey) return undefined;
 		return candidate;
 	}
@@ -1454,10 +1460,18 @@ export class SessionMaintenance {
 		const candidates =
 			precomputedCandidates ?? this.#getCompactionModelCandidates(this.#host.modelRegistry.getAvailable());
 		const telemetry = resolveTelemetry(this.#host.agent.telemetry, this.#host.sessionId());
+		let nativeCompactionFailure: { error: NativeCompactionError; provider: string } | undefined;
 
 		for (const candidate of candidates) {
 			const apiKey = await this.#host.modelRegistry.getApiKey(candidate, this.#host.sessionId());
 			if (!apiKey) continue;
+			if (
+				nativeCompactionFailure &&
+				(candidate.provider !== nativeCompactionFailure.provider ||
+					!shouldUseProviderNativeCompaction(candidate, preparation.settings))
+			) {
+				throw nativeCompactionFailure.error;
+			}
 
 			try {
 				return await compact(
@@ -1495,12 +1509,17 @@ export class SessionMaintenance {
 					},
 				);
 			} catch (error) {
-				if (!AIError.is(AIError.classify(error, candidate.api), AIError.Flag.AuthFailed)) {
-					throw error;
+				const id = AIError.classify(error instanceof NativeCompactionError ? error.cause : error, candidate.api);
+				if (AIError.is(id, AIError.Flag.AuthFailed)) continue;
+				if (error instanceof NativeCompactionError) {
+					nativeCompactionFailure ??= { error, provider: candidate.provider };
+					continue;
 				}
+				throw error;
 			}
 		}
 
+		if (nativeCompactionFailure) throw nativeCompactionFailure.error;
 		throw this.#buildCompactionAuthError();
 	}
 
@@ -2473,6 +2492,7 @@ export class SessionMaintenance {
 				const telemetry = resolveTelemetry(this.#host.agent.telemetry, this.#host.sessionId());
 				let compactResult: CompactionResult | undefined;
 				let lastError: unknown;
+				let nativeCompactionFailure: { error: NativeCompactionError; provider: string } | undefined;
 				codexCompaction = createCodexCompactionContext({
 					trigger: "auto",
 					reason: "context_limit",
@@ -2486,6 +2506,13 @@ export class SessionMaintenance {
 					const hasMoreCandidates = candidateIndex < candidates.length - 1;
 					const apiKey = await this.#host.modelRegistry.getApiKey(candidate, this.#host.sessionId());
 					if (!apiKey) continue;
+					if (
+						nativeCompactionFailure &&
+						(candidate.provider !== nativeCompactionFailure.provider ||
+							!shouldUseProviderNativeCompaction(candidate, preparation.settings))
+					) {
+						throw nativeCompactionFailure.error;
+					}
 
 					let attempt = 0;
 					while (true) {
@@ -2523,22 +2550,33 @@ export class SessionMaintenance {
 							}
 
 							const message = error instanceof Error ? error.message : String(error);
-							const id = AIError.classify(error, candidate.api);
+							const id = AIError.classify(
+								error instanceof NativeCompactionError ? error.cause : error,
+								candidate.api,
+							);
 							if (AIError.is(id, AIError.Flag.AuthFailed)) {
-								lastError = this.#buildCompactionAuthError();
+								if (!nativeCompactionFailure) lastError = this.#buildCompactionAuthError();
 								break;
 							}
 							if (AIError.is(id, AIError.Flag.Timeout)) {
+								const nativeFailure = error instanceof NativeCompactionError;
 								logger.warn(
-									hasMoreCandidates
-										? "Auto-compaction summarization timed out, trying next model"
-										: "Auto-compaction summarization timed out, not retrying same model",
+									nativeFailure
+										? "Provider-native auto-compaction timed out, preserving native failure"
+										: hasMoreCandidates
+											? "Auto-compaction summarization timed out, trying next model"
+											: "Auto-compaction summarization timed out, not retrying same model",
 									{
 										error: message,
 										model: `${candidate.provider}/${candidate.id}`,
 									},
 								);
-								lastError = error;
+								if (nativeFailure) {
+									nativeCompactionFailure ??= { error, provider: candidate.provider };
+									lastError = nativeCompactionFailure.error;
+								} else {
+									lastError = error;
+								}
 								break;
 							}
 
@@ -2550,7 +2588,12 @@ export class SessionMaintenance {
 									AIError.is(id, AIError.Flag.Transient) ||
 									AIError.is(id, AIError.Flag.UsageLimit));
 							if (!shouldRetry) {
-								lastError = error;
+								if (error instanceof NativeCompactionError) {
+									nativeCompactionFailure ??= { error, provider: candidate.provider };
+									lastError = nativeCompactionFailure.error;
+								} else {
+									lastError = error;
+								}
 								break;
 							}
 
@@ -2560,6 +2603,11 @@ export class SessionMaintenance {
 							// If retry delay is too long (>30s), try next candidate instead of waiting
 							const maxAcceptableDelayMs = 30_000;
 							if (delayMs > maxAcceptableDelayMs && hasMoreCandidates) {
+								if (error instanceof NativeCompactionError) {
+									nativeCompactionFailure ??= { error, provider: candidate.provider };
+									lastError = nativeCompactionFailure.error;
+									break;
+								}
 								logger.warn("Auto-compaction retry delay too long, trying next model", {
 									delayMs,
 									retryAfterMs,

package/src/session/session-manager.ts

@@ -35,6 +35,7 @@ import { type BuildSessionContextOptions, buildSessionContext, type SessionConte
 import {
 	type BranchSummaryEntry,
 	type CompactionEntry,
+	type CredentialPinEntry,
 	CURRENT_SESSION_VERSION,
 	type CustomEntry,
 	type CustomMessageEntry,
@@ -179,6 +180,7 @@ function isDraftOnlyMetadataEntry(entry: SessionEntry): boolean {
 		case "thinking_level_change":
 		case "service_tier_change":
 		case "mode_change":
+		case "credential_pin":
 			return true;
 		default:
 			return false;
@@ -1420,6 +1422,30 @@ export class SessionManager {
 		await this.#rewriteAtomically();
 	}
 
+	/** Persist this session's transcript as a newly identified OMP session. */
+	async persistCopy(
+		options?: { sessionDir?: string; suppressBreadcrumb?: boolean },
+		storage: SessionStorage = new FileSessionStorage(),
+	): Promise<SessionManager> {
+		const sessionDir = options?.sessionDir ?? SessionManager.getDefaultSessionDir(this.#cwd, undefined, storage);
+		const manager = new SessionManager(this.#cwd, sessionDir, true, storage);
+		manager.#suppressBreadcrumb = options?.suppressBreadcrumb === true;
+		manager.#resetToNewSession();
+		manager.#sessionName = this.#sessionName;
+		manager.#titleSource = this.#titleSource;
+		manager.#titleUpdatedAt = this.#titleUpdatedAt;
+		manager.#header.title = this.#sessionName;
+		manager.#header.titleSource = this.#titleSource;
+		manager.#additionalDirectories = [...this.#additionalDirectories];
+		manager.#header.additionalDirectories =
+			manager.#additionalDirectories.length > 0 ? [...manager.#additionalDirectories] : undefined;
+		manager.#entries = structuredClone(this.#entries);
+		manager.#index.rebuild(manager.#entries);
+		manager.#forceFileCreation = true;
+		await manager.#rewriteAtomically();
+		return manager;
+	}
+
 	/**
 	 * Stage a synchronous group of entry appends and publish the resulting full
 	 * journal with one atomic replace. A failed publish removes only the staged
@@ -2038,6 +2064,41 @@ export class SessionManager {
 		return [...names];
 	}
 
+	/** Append a credential pin recording which OAuth account served `provider`. */
+	appendCredentialPin(provider: string, hash: string): string {
+		const entry: CredentialPinEntry = {
+			type: "credential_pin",
+			...this.#freshEntryFields(),
+			provider,
+			hash,
+		};
+		this.#recordEntry(entry);
+		return entry.id;
+	}
+
+	/**
+	 * Latest credential pin per provider on the current branch (root → leaf),
+	 * with the effective last-use time of the pinned account.
+	 *
+	 * Pins are appended only when the serving account *changes*, so a long
+	 * session on one account carries a single old pin entry. Any assistant turn
+	 * for the same provider after that pin was necessarily served by the pinned
+	 * account, so its timestamp advances `lastUsedAt` — a resume seconds after
+	 * the last turn seeds a warm sticky instead of a stale one.
+	 */
+	getCredentialPins(): Map<string, { hash: string; lastUsedAt: number }> {
+		const pins = new Map<string, { hash: string; lastUsedAt: number }>();
+		for (const entry of this.getBranch()) {
+			if (entry.type === "credential_pin") {
+				pins.set(entry.provider, { hash: entry.hash, lastUsedAt: new Date(entry.timestamp).getTime() });
+			} else if (entry.type === "message" && entry.message.role === "assistant") {
+				const pin = pins.get(entry.message.provider);
+				if (pin) pin.lastUsedAt = Math.max(pin.lastUsedAt, entry.message.timestamp);
+			}
+		}
+		return pins;
+	}
+
 	getLeafId(): string | null {
 		return this.#index.leafId();
 	}

package/src/session/session-tools.ts

@@ -1,6 +1,6 @@
 import type { Agent, AgentTool } from "@oh-my-pi/pi-agent-core";
 import type { Model } from "@oh-my-pi/pi-ai";
-import { logger, prompt, stringProperty } from "@oh-my-pi/pi-utils";
+import { isRecord, logger, prompt, stringProperty } from "@oh-my-pi/pi-utils";
 import { reset as resetCapabilities } from "../capability";
 import type { ModelRegistry } from "../config/model-registry";
 import { formatModelString } from "../config/model-resolver";
@@ -159,6 +159,18 @@ export function projectMountedMCPXdevGuidance(routes: Iterable<MountedMCPToolRou
 
 const XDEV_MOUNT_NOTICE_MESSAGE_TYPE = "xdev-mount-notice";
 
+/**
+ * Structured payload persisted on each {@link XDEV_MOUNT_NOTICE_MESSAGE_TYPE}
+ * custom message. Lets a resumed session reconstruct which dynamic devices the
+ * model has already been told about, so reconnecting hosts do not re-announce
+ * (and re-splice a redundant developer message that busts the provider
+ * prompt-cache prefix).
+ */
+interface XdevMountNoticeDetails {
+	added: string[];
+	removed: string[];
+}
+
 /** Owns tool registration, presentation, prompt rebuilding, skills, and permissions. */
 export class SessionTools {
 	readonly #host: SessionToolsHost;
@@ -172,6 +184,14 @@ export class SessionTools {
 	#rpcHostToolNames = new Set<string>();
 	#xdev: XdevState | undefined;
 	#pendingXdevMountDelta: { added: Set<string>; removed: Set<string> } | undefined;
+	/**
+	 * Dynamic (`xd://`) devices the model has already been told are mounted.
+	 * Seeded lazily from persisted history on resume (see
+	 * {@link #ensureAnnouncedMountsSeeded}) and updated as notices are emitted, so
+	 * a host reconnect that re-mounts the same device does not re-announce it.
+	 */
+	#announcedMounts = new Set<string>();
+	#announcedMountsSeeded = false;
 	#presentationPinnedToolNames: ReadonlySet<string> | undefined;
 	#runtimeSelectedToolNames: ReadonlySet<string> | undefined;
 	#baseSystemPrompt: string[];
@@ -674,26 +694,108 @@ export class SessionTools {
 		this.#host.emitNotice("info", `xd://: ${parts.join("; ")}`, "xdev");
 	}
 
+	/**
+	 * Forget the announced-mount baseline for a replaced transcript. Called when
+	 * session history is swapped wholesale (`/new`, `switchSession`, `branch`): the
+	 * previous transcript's persisted notices no longer apply, so the next notice
+	 * re-seeds from the new history and a device reconnecting into it announces
+	 * again.
+	 *
+	 * The pending delta is deliberately preserved: it holds mounts that are still
+	 * live but not yet delivered to the model, and `branch()` does not rebuild the
+	 * base system prompt, so dropping it would leave the branched transcript
+	 * unaware of a still-mounted device that no later refresh would re-queue.
+	 */
+	resetAnnouncedMounts(): void {
+		this.#announcedMounts.clear();
+		this.#announcedMountsSeeded = false;
+	}
+
+	/**
+	 * Seed {@link #announcedMounts} from persisted mount notices the first time a
+	 * notice is consumed. On resume the in-memory mount set is rebuilt from
+	 * scratch, so without replaying history every already-announced dynamic device
+	 * would look freshly mounted and re-announce.
+	 */
+	#ensureAnnouncedMountsSeeded(): void {
+		if (this.#announcedMountsSeeded) return;
+		this.#announcedMountsSeeded = true;
+		for (const message of this.#host.agent.state.messages) {
+			if (message.role !== "custom" || message.customType !== XDEV_MOUNT_NOTICE_MESSAGE_TYPE) continue;
+			const details = message.details;
+			if (
+				isRecord(details) &&
+				Array.isArray(details.added) &&
+				details.added.every(name => typeof name === "string") &&
+				Array.isArray(details.removed) &&
+				details.removed.every(name => typeof name === "string")
+			) {
+				for (const name of details.added) this.#announcedMounts.add(name);
+				for (const name of details.removed) this.#announcedMounts.delete(name);
+				continue;
+			}
+
+			// Releases before structured notice details persisted only the rendered
+			// prompt. Replay its two stable inventory sections so the first resume
+			// after upgrading does not re-announce every dynamic device once.
+			if (typeof message.content !== "string") continue;
+			let section: "added" | "removed" | undefined;
+			for (const line of message.content.split("\n")) {
+				if (line === "These tools became available:") {
+					section = "added";
+					continue;
+				}
+				if (line.startsWith("No longer mounted")) {
+					section = "removed";
+					continue;
+				}
+				if (line === "Configured inline device docs:" || line === "</system-notice>") break;
+				if (line.startsWith("Read `xd://<tool>`")) {
+					section = undefined;
+					continue;
+				}
+				if (!section) continue;
+				const match = /^- xd:\/\/(\S+?)(?:\s+—|$)/.exec(line);
+				const name = match?.[1];
+				if (!name) continue;
+				if (section === "added") this.#announcedMounts.add(name);
+				else this.#announcedMounts.delete(name);
+			}
+		}
+	}
+
 	/** Consumes the hidden notice for unannounced `xd://` mount changes. */
-	takePendingXdevMountNotice(): CustomMessage | undefined {
+	takePendingXdevMountNotice(): CustomMessage<XdevMountNoticeDetails> | undefined {
 		const pending = this.#pendingXdevMountDelta;
 		if (!pending) return undefined;
 		this.#pendingXdevMountDelta = undefined;
+		this.#ensureAnnouncedMountsSeeded();
+		// Only announce a net change relative to what the model already knows (from
+		// this session and persisted history): a re-mount of an already-announced
+		// device — the common resume/reconnect case — and an unmount for a device
+		// it was never told about are both suppressed, keeping the provider prompt
+		// cache prefix byte-stable across resumes.
+		const addedNames = [...pending.added].filter(name => !this.#announcedMounts.has(name));
+		const removedNames = [...pending.removed].filter(name => this.#announcedMounts.has(name));
+		if (addedNames.length === 0 && removedNames.length === 0) return undefined;
 		const summaries = new Map(this.#xdev ? xdevEntries(this.#xdev).map(entry => [entry.name, entry.summary]) : []);
-		const added = [...pending.added].map(name => ({ name, summary: summaries.get(name) ?? "" }));
-		const removed = [...pending.removed].map(name => ({ name }));
+		const added = addedNames.map(name => ({ name, summary: summaries.get(name) ?? "" }));
+		const removed = removedNames.map(name => ({ name }));
 		const docs = this.#xdev
 			? xdevDocsFor(
 					this.#xdev,
-					pending.added,
+					new Set(addedNames),
 					this.#host.settings.get("tools.xdevDocs"),
 					this.#host.settings.get("tools.xdevInlineDevices"),
 				)
 			: "";
+		for (const name of addedNames) this.#announcedMounts.add(name);
+		for (const name of removedNames) this.#announcedMounts.delete(name);
 		return {
 			role: "custom",
 			customType: XDEV_MOUNT_NOTICE_MESSAGE_TYPE,
 			content: prompt.render(xdevMountNoticePrompt, { added, removed, docs }),
+			details: { added: addedNames, removed: removedNames },
 			attribution: "agent",
 			display: false,
 			timestamp: Date.now(),

package/src/session/settings-stream-fn.ts

@@ -58,6 +58,7 @@ export function createSettingsAwareStreamFn(settings: Settings, base: StreamFn =
 			textVerbosity: streamOptions?.textVerbosity ?? textVerbosity,
 			streamFirstEventTimeoutMs: streamOptions?.streamFirstEventTimeoutMs ?? streamFirstEventTimeoutMs,
 			streamIdleTimeoutMs: streamOptions?.streamIdleTimeoutMs ?? streamIdleTimeoutMs,
+			maxRetryDelayMs: streamOptions?.maxRetryDelayMs ?? settings.get("retry.maxDelayMs"),
 			maxInFlightRequests: validateProviderMaxInFlightRequests(
 				streamOptions?.maxInFlightRequests ?? settings.get("providers.maxInFlightRequests"),
 			),