@vanillagreen/pi-claude-bridge 1.0.0

package/src/index.ts

@@ -0,0 +1,1225 @@
+import { calculateCost, getModels, type AssistantMessage, type AssistantMessageEventStream, type Context, type Model, type SimpleStreamOptions, type Tool } from "@mariozechner/pi-ai";
+import * as piAi from "@mariozechner/pi-ai";
+import { type ExtensionAPI, type ExtensionUIContext } from "@mariozechner/pi-coding-agent";
+import { createSdkMcpServer, query, type EffortLevel, type SDKMessage, type SDKUserMessage, type SettingSource } from "@anthropic-ai/claude-agent-sdk";
+import type { Base64ImageSource, ContentBlockParam, MessageParam } from "@anthropic-ai/sdk/resources";
+import { createSession, deleteSession, repairToolPairing } from "cc-session-io";
+import { accessSync, appendFileSync, constants as fsConstants, mkdirSync, realpathSync, statSync } from "fs";
+import { homedir } from "os";
+import { delimiter, dirname, join } from "path";
+import { PROVIDER_ID, messageContentToText, convertPiMessages } from "./convert.js";
+import { buildModels } from "./models.js";
+import { MCP_SERVER_NAME, MCP_TOOL_PREFIX, extractSkillsBlock } from "./skills.js";
+import { verifyWrittenSession as _verifyWrittenSession } from "./session-verify.js";
+import { extractAllToolResults as _extractAllToolResults, type McpResult } from "./extract-tool-results.js";
+import { QueryContext, ctx, stackDepth, pushContext, popContext } from "./query-state.js";
+import { loadConfig } from "./config.js";
+import { extractAgentsAppend } from "./agents-md.js";
+import { buildPromptContextAppend } from "./prompt-context.js";
+import { jsonSchemaToZodShape } from "./typebox-to-zod.js";
+
+// Compat (#2): use factory if available (pi-ai ≥0.66), else fall back to constructor (gsd-pi etc.)
+const _piAi = piAi as any;
+const newAssistantMessageEventStream: () => AssistantMessageEventStream =
+	typeof _piAi.createAssistantMessageEventStream === "function"
+		? _piAi.createAssistantMessageEventStream
+		: () => new _piAi.AssistantMessageEventStream();
+
+// --- Debug logging ---
+// CLAUDE_BRIDGE_DEBUG=1 enables debug logging to ~/.pi/agent/claude-bridge.log
+
+const DEBUG = process.env.CLAUDE_BRIDGE_DEBUG === "1";
+const DEBUG_LOG_PATH = process.env.CLAUDE_BRIDGE_DEBUG_PATH || join(homedir(), ".pi", "agent", "claude-bridge.log");
+const DIAG_LOG_PATH = join(homedir(), ".pi", "agent", "claude-bridge-diag.log");
+
+// Ensure log directories exist when debug is enabled
+if (DEBUG) {
+	try {
+		mkdirSync(dirname(DEBUG_LOG_PATH), { recursive: true });
+		mkdirSync(dirname(DIAG_LOG_PATH), { recursive: true });
+	} catch {
+		// If directory creation fails, debug functions will throw on first use
+	}
+}
+
+// Unique per module evaluation — confirms whether subagents share module state
+const moduleInstanceId = Math.random().toString(36).slice(2, 8);
+
+function debug(...args: unknown[]) {
+	if (!DEBUG) return;
+	const ts = new Date().toISOString();
+	const fmt = (a: unknown): string => {
+		if (typeof a === "string") return a;
+		if (a instanceof Error) return `${a.name}: ${a.message}${a.stack ? "\n" + a.stack : ""}`;
+		return JSON.stringify(a);
+	};
+	const msg = args.map(fmt).join(" ");
+	appendFileSync(DEBUG_LOG_PATH, `[${ts}] [${moduleInstanceId}] ${msg}\n`);
+}
+
+function executableFromPath(name: string): string | undefined {
+	const paths = (process.env.PATH ?? "").split(delimiter).filter(Boolean);
+	for (const dir of paths) {
+		const candidate = join(dir, name);
+		try {
+			accessSync(candidate, fsConstants.X_OK);
+			return candidate;
+		} catch {
+			// keep searching
+		}
+	}
+	return undefined;
+}
+
+function resolveClaudeExecutable(configured?: string): string | undefined {
+	const trimmed = configured?.trim();
+	if (trimmed) return trimmed;
+	return executableFromPath("claude") ?? executableFromPath("claude-code");
+}
+
+// Per-query CLI debug capture. When CLAUDE_BRIDGE_DEBUG=1, ask the Claude Code
+// CLI subprocess to write its own debug log to a file we choose, and also
+// forward its stderr into our debug stream. Drops straight into the real SDK's
+// Options — see @anthropic-ai/claude-agent-sdk sdk.d.ts:1245 (debug, debugFile,
+// stderr). Without this, CC's internal view of the world is invisible to us
+// and "No conversation found" / empty-error reports are unactionable.
+let nextCliDebugSeq = 1;
+function makeCliDebugOptions(tag: string): { debug?: boolean; debugFile?: string; stderr?: (data: string) => void } {
+	if (!DEBUG) return {};
+	const seq = nextCliDebugSeq++;
+	const ts = new Date().toISOString().replace(/[:.]/g, "-");
+	const logDir = join(dirname(DEBUG_LOG_PATH), "cc-cli-logs");
+	try { mkdirSync(logDir, { recursive: true }); } catch { /* ignore */ }
+	const debugFile = join(logDir, `${ts}-${tag}-${seq}.log`);
+	debug(`cli-debug: ${tag} #${seq} → ${debugFile}`);
+	return {
+		debug: true,
+		debugFile,
+		stderr: (data: string) => {
+			for (const line of data.split(/\r?\n/)) {
+				if (line) debug(`[cli-stderr ${tag}#${seq}] ${line}`);
+			}
+		},
+	};
+}
+
+/** Unconditional diagnostic dump — for "should never happen" paths */
+function diagDump(label: string, data: Record<string, unknown>) {
+	const ts = new Date().toISOString();
+	const entry = { ts, moduleInstanceId, label, ...data };
+	appendFileSync(DIAG_LOG_PATH, JSON.stringify(entry) + "\n");
+	debug(`DIAG: ${label} (see ${DIAG_LOG_PATH})`);
+}
+
+// --- Constants ---
+
+// Global key to prevent re-registration of the provider across module reloads.
+//
+// Extensions like pi-subagents spawn a subagent and it loads this module
+// again. Without this guard, the subagent's call to registerProvider() would
+// overwrite the parent's `streamSimple` function reference in the shared
+// ModelRegistry. When the parent later delivers a tool result, it would call
+// the subagent's `streamSimple` (which has empty state) instead of its own.
+//
+// By storing the active streamSimple in a Symbol.for() global (shared across all
+// module instances), we ensure only the FIRST instance to register takes effect.
+// Subsequent instances wrap the stored function instead of overwriting it.
+//
+// On session_shutdown (including /reload), clearSession() resets this so a fresh
+// registration can occur for the next session.
+const ACTIVE_STREAM_SIMPLE_KEY = Symbol.for("claude-bridge:activeStreamSimple");
+
+const SDK_TO_PI_TOOL_NAME: Record<string, string> = {
+	read: "read", write: "write", edit: "edit", bash: "bash",
+};
+
+// MODELS is buildModels(getModels("anthropic")) — projection kept in models.js.
+const MODELS = buildModels(getModels("anthropic"));
+
+// Disable Claude Code built-ins in the provider path. Pi owns tool execution;
+// Claude reaches Pi tools through the bridged MCP server instead.
+const DISALLOWED_BUILTIN_TOOLS = [
+	"Read", "Write", "Edit", "Glob", "Grep", "Bash", "Agent",
+	"NotebookEdit", "EnterWorktree", "ExitWorktree",
+	"CronCreate", "CronDelete", "TeamCreate", "TeamDelete",
+	"WebFetch", "WebSearch",
+	"AskUserQuestion", "EnterPlanMode", "ExitPlanMode",
+	"ToolSearch", "ScheduleWakeup",
+];
+
+// --- Session persistence ---
+
+interface SessionState {
+	sessionId: string;
+	cursor: number;
+	cwd: string;
+	// Force the next syncSharedSession call down the REBUILD path. Set when
+	// pi has mutated its messages array out from under us (compact, tree
+	// navigation) or after an abort left the JSONL in an indeterminate state.
+	// REBUILD wipes and rewrites the file to match pi's current history.
+	needsRebuild?: boolean;
+	// Set ONLY after an abort. The killed CC subprocess may still be flushing
+	// a late "[Request interrupted by user]" record to the session JSONL.
+	// Reusing the same sessionId/path would race that orphan write into our
+	// fresh file and break CC's parent-uuid chain on the next resume. When
+	// this flag is set, REBUILD takes a fresh UUID and skips deleteSession
+	// so the orphan writes land on a dead inode. Compact/tree do NOT set
+	// this — there's no concurrent CC writer during those events, so
+	// in-place rebuild (preserve UUID, deleteSession + createSession) is safe.
+	forceRotate?: boolean;
+}
+
+let sharedSession: SessionState | null = null;
+
+// Convert pi messages to Anthropic API format for session import.
+// Lossy: non-Anthropic thinking blocks are dropped (no valid signature), and only
+// text/image/toolCall block types are handled. If all blocks in an assistant message
+// are filtered, the message is dropped — which can create invalid sequences (e.g.
+// two user messages in a row, or tool_result without preceding tool_use).
+function convertAndImportMessages(
+	session: ReturnType<typeof createSession>,
+	messages: Context["messages"],
+	customToolNameToSdk?: Map<string, string>,
+): void {
+	const { anthropicMessages, sanitizedIds } = convertPiMessages(messages, customToolNameToSdk);
+
+	debug(`convertAndImportMessages: ${messages.length} pi msgs → ${anthropicMessages.length} anthropic msgs`);
+	debug(`convertAndImportMessages: imported roles:`, anthropicMessages.map((m, i) => {
+		const c = m.content;
+		if (typeof c === "string") return `[${i}]${m.role}:text`;
+		if (Array.isArray(c)) return `[${i}]${m.role}:${(c).map((b) => b.type).join("+")}`;
+		return `[${i}]${m.role}:?`;
+	}).join(" "));
+	if (sanitizedIds.size > 0) {
+		debug(`convertAndImportMessages: sanitized ${sanitizedIds.size} tool IDs:`,
+			[...sanitizedIds.entries()].map(([orig, clean]) => orig === clean ? orig : `${orig}→${clean}`).join(", "));
+	}
+	// Pre-repair for debug logging; importMessages also repairs internally (idempotent).
+	const repaired = repairToolPairing(anthropicMessages);
+	if (repaired.length !== anthropicMessages.length) {
+		debug(`convertAndImportMessages: repairToolPairing ${anthropicMessages.length} → ${repaired.length} msgs`);
+	}
+	if (repaired.length) session.importMessages(repaired);
+}
+
+// Pi doesn't pass tool results directly — it appends them to the context and calls
+// the provider again. Thin wrapper over extract-tool-results.js that adds per-turn
+// debug logging at the extraction boundary.
+function extractAllToolResults(context: Context): McpResult[] {
+	const { results, stopIdx } = _extractAllToolResults(context.messages as unknown as Array<{ role: string; [key: string]: unknown }>);
+	debug(`extractAllToolResults: ${results.length} results from ${context.messages.length} msgs, stopped at index ${stopIdx}`);
+	debug(`extractAllToolResults: all msg roles:`, context.messages.map((m, i) => `[${i}]${m.role}`).join(" "));
+	for (let r = 0; r < results.length; r++) {
+		debug(`extractAllToolResults: result[${r}] id=${results[r].toolCallId}${results[r].isError ? " ERROR" : ""} preview:`, JSON.stringify(results[r].content).slice(0, 150));
+	}
+	return results;
+}
+
+/** Extract the last user message from context as a prompt string. Returns null if last message is not a user message. */
+function extractUserPrompt(messages: Context["messages"]): string | null {
+	const last = messages[messages.length - 1];
+	if (!last || last.role !== "user") return null;
+	if (typeof last.content === "string") return last.content;
+	return messageContentToText(last.content) || "";
+}
+
+/** Extract the last user message as ContentBlockParam[] (preserving images).
+ *  Returns null if no images — caller should fall back to string prompt. */
+function extractUserPromptBlocks(messages: Context["messages"]): ContentBlockParam[] | null {
+	const last = messages[messages.length - 1];
+	if (!last || last.role !== "user") return null;
+	if (typeof last.content === "string") {
+		debug(`extractUserPromptBlocks: content is string (length=${last.content.length})`);
+		return null;
+	}
+	if (!Array.isArray(last.content)) {
+		debug(`extractUserPromptBlocks: content is ${typeof last.content}`);
+		return null;
+	}
+	debug(`extractUserPromptBlocks: ${last.content.length} blocks, types=${last.content.map((b: any) => b.type).join(",")}`);
+	let hasImage = false;
+	const blocks: ContentBlockParam[] = [];
+	for (const block of last.content) {
+		if (block.type === "text" && block.text) {
+			blocks.push({ type: "text", text: block.text });
+		} else if (block.type === "image") {
+			debug(`image block: mimeType=${(block as any).mimeType}, data length=${((block as any).data ?? "").length}, keys=${Object.keys(block).join(",")}`);
+			if (!(block as any).data || !(block as any).mimeType) {
+				debug(`image block missing data or mimeType, skipping`);
+				continue;
+			}
+			hasImage = true;
+			blocks.push({
+				type: "image",
+				source: { type: "base64", media_type: block.mimeType as Base64ImageSource["media_type"], data: block.data },
+			});
+		}
+	}
+	return hasImage ? blocks : null;
+}
+
+async function* wrapPromptStream(blocks: ContentBlockParam[]): AsyncIterable<SDKUserMessage> {
+	yield {
+		type: "user",
+		message: { role: "user", content: blocks } as MessageParam,
+		parent_tool_use_id: null,
+	};
+}
+
+
+interface SyncResult {
+	sessionId: string | null;
+}
+
+/**
+ * Ensure the shared session has all messages up to (but not including) the last user message.
+ * Returns session ID to resume from, or null if no resume needed.
+ */
+// Read the session file we just wrote and sanity-check it. Warns instead of
+// throwing — CC may be more tolerant than our checks, so a false positive
+// shouldn't block the user. Pure logic is in session-verify.js; this wrapper
+// fans each warning out to debug log + piUI notify + diagDump.
+function verifyWrittenSession(
+	jsonlPath: string,
+	expectedSessionId: string,
+	expectedRecordCount: number,
+	cwd: string,
+): void {
+	const warnings = _verifyWrittenSession(jsonlPath, expectedSessionId, expectedRecordCount);
+	for (const msg of warnings) {
+		debug(`WARNING session verify: ${msg}`);
+		piUI?.notify(
+			`Session file issue: ${msg}\n` +
+			`cwd=${cwd} realpath=${safeRealpath(cwd)} CLAUDE_CONFIG_DIR=${process.env.CLAUDE_CONFIG_DIR ?? "(unset)"}\n` +
+			`Please copy and paste this message into a new issue at https://github.com/elidickinson/pi-claude-bridge/issues/new` +
+			(DEBUG ? ` and attach ${DEBUG_LOG_PATH}` : ` (rerun with CLAUDE_BRIDGE_DEBUG=1 to capture a debug log)`),
+			"warning",
+		);
+		diagDump("session_verify_fail", { msg, jsonlPath, cwd, realpath: safeRealpath(cwd), claudeConfigDir: process.env.CLAUDE_CONFIG_DIR ?? null });
+	}
+}
+
+function safeRealpath(p: string): string {
+	try { return realpathSync(p); } catch (e) { return `<failed: ${(e as Error).message}>`; }
+}
+
+// Diagnostic snapshot of where a session file was just written. Catches the
+// class of bugs where pi writes to ~/.claude/projects/<X> but CC SDK reads
+// from ~/.claude/projects/<Y> (symlinks, CLAUDE_CONFIG_DIR, hash mismatch).
+function debugSessionPaths(label: string, cwd: string, jsonlPath: string): void {
+	const realCwd = safeRealpath(cwd);
+	let fileSize: number | null = null;
+	let fileExists = false;
+	try {
+		const st = statSync(jsonlPath);
+		fileExists = true;
+		fileSize = st.size;
+	} catch { /* file may not exist yet */ }
+	debug(`${label}: cwd=${cwd}`);
+	if (realCwd !== cwd) debug(`${label}: realpath(cwd)=${realCwd} (DIFFERS — symlink-resolved path is what CC SDK uses)`);
+	debug(`${label}: jsonlPath=${jsonlPath}`);
+	debug(`${label}: fileExists=${fileExists}${fileSize != null ? ` size=${fileSize}` : ""}`);
+	debug(`${label}: env.CLAUDE_CONFIG_DIR=${process.env.CLAUDE_CONFIG_DIR ?? "(unset)"} HOME=${process.env.HOME ?? "(unset)"}`);
+}
+
+// Two semantic paths:
+//   REUSE — pi's history is in sync with the existing sharedSession (or drifted
+//     only by the trailing final-assistant message that pi appends after
+//     streamSimple returns, which CC's own persisted session already has).
+//     Returns the existing sessionId. Keeps CC's prompt cache warm.
+//   REBUILD — no session yet, or pi's history has diverged (non-trailing
+//     missed messages, e.g. another provider took a turn). Wipes the existing
+//     session file (if any) and writes a fresh one containing all prior
+//     messages, reusing the same sessionId across rebuilds so UUIDs stay
+//     stable for the lifetime of pi's session.
+//
+// Why a full rebuild rather than patching:
+//   Injecting deltas into an existing session creates a branch that CC's
+//   --resume doesn't follow (documented attempt prior to this). A complete
+//   overwrite at the same path is simpler and correct.
+//
+// Why reuse the sessionId across rebuilds:
+//   CC re-reads the JSONL on every --resume call — no in-process UUID
+//   caching. Validated in tests/exp-session-clear.mjs, including the case
+//   where CC had appended its own tool_use/tool_result records between
+//   rebuilds. Preserving the UUID means stable log correlation across
+//   provider switches and no orphaned session files.
+//
+// Log strings still say "Case 1/2/3/4" so existing diagnostics (int-cache.sh,
+// int-session-resume.mjs) keep grepping the same anchors.
+function syncSharedSession(
+	messages: Context["messages"],
+	cwd: string,
+	customToolNameToSdk?: Map<string, string>,
+	modelId?: string,
+): SyncResult {
+	const priorMessages = messages.slice(0, -1); // everything before the new user prompt
+
+	// REUSE path
+	if (sharedSession && !sharedSession.needsRebuild) {
+		const missed = priorMessages.slice(sharedSession.cursor);
+		const trailingAssistantOnly =
+			missed.length === 1 && (missed[0] as { role?: string }).role === "assistant";
+		if (missed.length === 0 || trailingAssistantOnly) {
+			if (trailingAssistantOnly) {
+				sharedSession = { ...sharedSession, cursor: priorMessages.length, cwd };
+			}
+			debug(`Case 3: ${trailingAssistantOnly ? "advanced cursor past trailing assistant, " : ""}resuming session ${sharedSession.sessionId.slice(0, 8)}, cursor=${sharedSession.cursor}`);
+			debug(`syncResult: path=reuse sessionId=${sharedSession.sessionId} cursor=${sharedSession.cursor}`);
+			return { sessionId: sharedSession.sessionId };
+		}
+	}
+
+	// REBUILD path
+	if (priorMessages.length === 0) {
+		debug(`Case 1: clean start, ${messages.length} total messages`);
+		debug(`syncResult: path=clean-start`);
+		return { sessionId: null };
+	}
+	const previousSessionId = sharedSession?.sessionId;
+	const previousCursor = sharedSession?.cursor ?? 0;
+	// preserveId: rebuild in place (deleteSession + createSession with the
+	// existing UUID), so prompt-cache UUIDs stay stable for log correlation
+	// and for any tools that key off them. Skipped only when there's a
+	// concurrent writer we shouldn't race — see forceRotate docs above.
+	const preserveId = previousSessionId !== undefined && !sharedSession?.forceRotate;
+	if (preserveId) {
+		// Wipe prior jsonl + companion dir (no-op if nothing to wipe).
+		deleteSession(previousSessionId!, cwd, process.env.CLAUDE_CONFIG_DIR);
+	}
+	const session = createSession({
+		projectPath: cwd,
+		claudeDir: process.env.CLAUDE_CONFIG_DIR,
+		...(preserveId ? { sessionId: previousSessionId } : {}),
+		...(modelId ? { model: modelId } : {}),
+	});
+	convertAndImportMessages(session, priorMessages, customToolNameToSdk);
+	session.save();
+	verifyWrittenSession(session.jsonlPath, session.sessionId, session.messages.length, cwd);
+	sharedSession = { sessionId: session.sessionId, cursor: priorMessages.length, cwd };
+	if (previousSessionId === undefined) {
+		debug(`Case 2: first turn with ${priorMessages.length} prior messages → session ${session.sessionId.slice(0, 8)}, ${session.messages.length} records`);
+	} else if (preserveId) {
+		const missedCount = priorMessages.length - previousCursor;
+		debug(`Case 4: ${missedCount} missed messages, ${priorMessages.length} total → rewrote session ${session.sessionId.slice(0, 8)} (same id), ${session.messages.length} records`);
+	} else {
+		debug(`Case 4 post-abort: ${priorMessages.length} total → new session ${session.sessionId.slice(0, 8)} (was ${previousSessionId.slice(0, 8)}, rotated to avoid race with orphan writer), ${session.messages.length} records`);
+	}
+	debugSessionPaths(`${session.sessionId.slice(0, 8)}`, cwd, session.jsonlPath);
+	debug(`syncResult: path=rebuild sessionId=${session.sessionId} priors=${priorMessages.length} ${previousSessionId === undefined ? "first" : preserveId ? "preserved" : "rotated-post-abort"}`);
+	return { sessionId: session.sessionId };
+}
+
+// --- Provider helpers: tool name mapping ---
+
+function mapToolName(name: string, customToolNameToPi?: Map<string, string>): string {
+	const normalized = name.toLowerCase();
+	const builtin = SDK_TO_PI_TOOL_NAME[normalized];
+	if (builtin) return builtin;
+	if (customToolNameToPi) {
+		const mapped = customToolNameToPi.get(name) ?? customToolNameToPi.get(normalized);
+		if (mapped) return mapped;
+	}
+	if (normalized.startsWith(MCP_TOOL_PREFIX)) return name.slice(MCP_TOOL_PREFIX.length);
+	return name;
+}
+
+// Renames for Claude Code SDK param names that differ from pi's native names.
+// Keys not listed here pass through unchanged, so new pi params work automatically.
+const SDK_KEY_RENAMES: Record<string, Record<string, string>> = {
+	read:  { file_path: "path" },
+	write: { file_path: "path" },
+	edit:  { file_path: "path", old_string: "oldText", new_string: "newText", old_text: "oldText", new_text: "newText" },
+};
+
+// Maps SDK tool args to pi tool args via key renaming + pass-through.
+// Pi's own prepareArguments hooks handle any structural transforms (e.g. edit oldText/newText → edits[]).
+function mapToolArgs(
+	toolName: string, args: Record<string, unknown> | undefined,
+): Record<string, unknown> {
+	const input = args ?? {};
+	const renames = SDK_KEY_RENAMES[toolName.toLowerCase()];
+	const result: Record<string, unknown> = {};
+	for (const [key, value] of Object.entries(input)) {
+		const piKey = renames?.[key] ?? key;
+		if (!(piKey in result)) result[piKey] = value; // first alias wins
+	}
+	// Pi bash has no default timeout; add a safety default
+	if (toolName.toLowerCase() === "bash" && result.timeout == null) {
+		result.timeout = 120;
+	}
+	return result;
+}
+
+// --- Provider helpers: tool resolution ---
+
+// --- Provider helpers: tool bridge ---
+
+// --- Query state ---
+// QueryContext + context stack live in query-state.js so tests can import
+// them without activating the extension. `ctx()`, `pushContext()`, `popContext()`
+// are imported at the top of this file.
+
+// Global (not query state):
+let piUI: ExtensionUIContext | null = null;
+
+function resolveMcpTools(context: Context, excludeToolName?: string): {
+	mcpTools: Tool[];
+	customToolNameToSdk: Map<string, string>;
+	customToolNameToPi: Map<string, string>;
+} {
+	const mcpTools: Tool[] = [];
+	const customToolNameToSdk = new Map<string, string>();
+	const customToolNameToPi = new Map<string, string>();
+
+	if (!context.tools) return { mcpTools, customToolNameToSdk, customToolNameToPi };
+
+	for (const tool of context.tools) {
+		if (tool.name === excludeToolName) continue;
+		const sdkName = `${MCP_TOOL_PREFIX}${tool.name}`;
+		mcpTools.push(tool);
+		customToolNameToSdk.set(tool.name, sdkName);
+		customToolNameToSdk.set(tool.name.toLowerCase(), sdkName);
+		customToolNameToPi.set(sdkName, tool.name);
+		customToolNameToPi.set(sdkName.toLowerCase(), tool.name);
+	}
+
+	return { mcpTools, customToolNameToSdk, customToolNameToPi };
+}
+
+// Creates an MCP server that bridges pi tools to the SDK. Each tool handler
+// blocks on a Promise until pi delivers the tool result via streamSimple.
+// Handlers are assigned toolCallIds from turnToolCallIds (populated when the SDK
+// emits tool_use blocks). Results are matched by ID, not position.
+// Handlers close over the captured `queryCtx`, ensuring they operate on the
+// correct query's state even across pushContext/popContext calls.
+function buildMcpServers(tools: Tool[], queryCtx: QueryContext): Record<string, ReturnType<typeof createSdkMcpServer>> | undefined {
+	if (!tools.length) return undefined;
+	const mcpTools = tools.map((tool) => ({
+		name: tool.name,
+		description: tool.description,
+		inputSchema: jsonSchemaToZodShape(tool.parameters),
+		handler: async () => {
+			const toolCallId = queryCtx.turnToolCallIds[queryCtx.nextHandlerIdx++];
+			if (!toolCallId) debug(`WARNING: mcp handler ${tool.name} has no toolCallId (idx=${queryCtx.nextHandlerIdx - 1}, available=${queryCtx.turnToolCallIds.length})`);
+			if (toolCallId && queryCtx.pendingResults.has(toolCallId)) {
+				const result = queryCtx.pendingResults.get(toolCallId)!;
+				queryCtx.pendingResults.delete(toolCallId);
+				debug(`mcp handler: ${tool.name} [${toolCallId}] → resolved from queue (${queryCtx.pendingResults.size} remaining)`);
+				return result;
+			}
+			debug(`mcp handler: ${tool.name} [${toolCallId}] → waiting`);
+			return new Promise<McpResult>((resolve) => {
+				queryCtx.pendingToolCalls.set(toolCallId, { toolName: tool.name, resolve });
+			});
+		},
+	}));
+	const server = createSdkMcpServer({ name: MCP_SERVER_NAME, version: "1.0.0", tools: mcpTools });
+	return { [MCP_SERVER_NAME]: server };
+}
+
+// --- Usage helpers ---
+
+function updateUsage(output: AssistantMessage, usage: Record<string, number | undefined>, model: Model<any>): void {
+	if (usage.input_tokens != null) output.usage.input = usage.input_tokens;
+	if (usage.output_tokens != null) output.usage.output = usage.output_tokens;
+	if (usage.cache_read_input_tokens != null) output.usage.cacheRead = usage.cache_read_input_tokens;
+	if (usage.cache_creation_input_tokens != null) output.usage.cacheWrite = usage.cache_creation_input_tokens;
+	output.usage.totalTokens = output.usage.input + output.usage.output + output.usage.cacheRead + output.usage.cacheWrite;
+	calculateCost(model, output.usage);
+	const promptTokens = output.usage.input + output.usage.cacheRead + output.usage.cacheWrite;
+	const cachePct = promptTokens > 0 ? Math.round(output.usage.cacheRead / promptTokens * 100) : 0;
+	debug(`usage: in=${output.usage.input} out=${output.usage.output} cacheRead=${output.usage.cacheRead} cacheWrite=${output.usage.cacheWrite} total=${output.usage.totalTokens} cachePct=${cachePct}% model=${model.id}`);
+}
+
+// --- Effort level mapping ---
+// Pi reasoning levels → CC SDK effort levels
+
+const REASONING_TO_EFFORT: Record<string, EffortLevel> = {
+	minimal: "low", low: "low", medium: "medium", high: "high", xhigh: "max",
+};
+
+// --- Provider helpers: misc ---
+
+function mapStopReason(reason: string | undefined): "stop" | "length" | "toolUse" {
+	switch (reason) {
+		case "tool_use": return "toolUse";
+		case "max_tokens": return "length";
+		case "end_turn": default: return "stop";
+	}
+}
+
+function parsePartialJson(input: string, fallback: Record<string, unknown>): Record<string, unknown> {
+	if (!input) return fallback;
+	try { return JSON.parse(input); } catch { return fallback; }
+}
+
+
+// --- Provider: streaming function ---
+//
+// Push-based streaming with MCP tool bridge:
+// 1. streamSimple starts a query() and kicks off consumeQuery() in background
+// 2. consumeQuery() iterates the SDK generator, pushing events to currentPiStream
+// 3. On tool_use: ends the current pi stream, nulls it out. The MCP handler
+//    blocks the generator naturally — no events arrive until resolved.
+// 4. Pi executes the tool, calls streamSimple again. We swap in the new stream,
+//    resolve the MCP handler, and the generator unblocks — events flow to new stream.
+//
+// Note: resetTurnState clears turnSawStreamEvent while the generator may still
+// have queued messages from the previous turn. This is safe because step 3 nulls
+// currentPiStream, so any leftover messages hit the `!ctx().currentPiStream` guard
+// in consumeQuery and are skipped before resetTurnState runs.
+
+function ensureTurnStarted(): void {
+	if (!ctx().turnStarted && ctx().currentPiStream && ctx().turnOutput) {
+		ctx().currentPiStream!.push({ type: "start", partial: ctx().turnOutput });
+		ctx().turnStarted = true;
+	}
+}
+
+function finalizeCurrentStream(stopReason?: string): void {
+	if (!ctx().currentPiStream || !ctx().turnOutput) return;
+	debug(`provider: finalizeCurrentStream called, stopReason=${stopReason}, turnOutput=${JSON.stringify({stopReason: ctx().turnOutput!.stopReason, error: ctx().turnOutput!.errorMessage})}`);
+	if (!ctx().turnStarted) ensureTurnStarted();
+	const reason = stopReason === "length" ? "length" : "stop";
+	ctx().currentPiStream!.push({ type: "done", reason, message: ctx().turnOutput });
+	ctx().currentPiStream!.end();
+	ctx().currentPiStream = null;
+}
+
+/** Maps Anthropic stream events to pi stream events (text, thinking, toolcall).
+ *  On message_stop with tool_use: ends currentPiStream so pi can execute the tool. */
+function processStreamEvent(
+	message: SDKMessage,
+	customToolNameToPi: Map<string, string>,
+	model: Model<any>,
+): void {
+	const c = ctx();
+	if (!c.currentPiStream || !c.turnOutput) return;
+	c.turnSawStreamEvent = true;
+	const event = (message as SDKMessage & { event: any }).event;
+
+	if (event?.type === "message_start") {
+		c.turnToolCallIds = [];
+		c.nextHandlerIdx = 0;
+		if (event.message?.usage) updateUsage(c.turnOutput, event.message.usage, model);
+		return;
+	}
+
+	if (event?.type === "content_block_start") {
+		ensureTurnStarted();
+		if (event.content_block?.type === "text") {
+			c.turnBlocks.push({ type: "text", text: "", index: event.index });
+			c.currentPiStream!.push({ type: "text_start", contentIndex: c.turnBlocks.length - 1, partial: c.turnOutput });
+		} else if (event.content_block?.type === "thinking") {
+			c.turnBlocks.push({ type: "thinking", thinking: "", thinkingSignature: "", index: event.index });
+			c.currentPiStream!.push({ type: "thinking_start", contentIndex: c.turnBlocks.length - 1, partial: c.turnOutput });
+		} else if (event.content_block?.type === "tool_use") {
+			c.turnSawToolCall = true;
+			c.turnToolCallIds.push(event.content_block.id);
+			c.turnBlocks.push({
+				type: "toolCall", id: event.content_block.id,
+				name: mapToolName(event.content_block.name, customToolNameToPi),
+				arguments: (event.content_block.input as Record<string, unknown>) ?? {},
+				partialJson: "", index: event.index,
+			});
+			c.currentPiStream!.push({ type: "toolcall_start", contentIndex: c.turnBlocks.length - 1, partial: c.turnOutput });
+		} else {
+			debug("processStreamEvent: unhandled content_block_start type", event.content_block?.type);
+		}
+		return;
+	}
+
+	if (event?.type === "content_block_delta") {
+		const index = c.turnBlocks.findIndex((b: any) => b.index === event.index);
+		const block = c.turnBlocks[index];
+		if (!block) return;
+		if (event.delta?.type === "text_delta" && block.type === "text") {
+			block.text += event.delta.text;
+			c.currentPiStream!.push({ type: "text_delta", contentIndex: index, delta: event.delta.text, partial: c.turnOutput });
+		} else if (event.delta?.type === "thinking_delta" && block.type === "thinking") {
+			block.thinking += event.delta.thinking;
+			c.currentPiStream!.push({ type: "thinking_delta", contentIndex: index, delta: event.delta.thinking, partial: c.turnOutput });
+		} else if (event.delta?.type === "input_json_delta" && block.type === "toolCall") {
+			block.partialJson += event.delta.partial_json;
+			block.arguments = parsePartialJson(block.partialJson, block.arguments);
+			c.currentPiStream!.push({ type: "toolcall_delta", contentIndex: index, delta: event.delta.partial_json, partial: c.turnOutput });
+		} else if (event.delta?.type === "signature_delta" && block.type === "thinking") {
+			block.thinkingSignature = (block.thinkingSignature ?? "") + event.delta.signature;
+		} else {
+			debug("processStreamEvent: unhandled content_block_delta type", event.delta?.type);
+		}
+		return;
+	}
+
+	if (event?.type === "content_block_stop") {
+		const index = c.turnBlocks.findIndex((b: any) => b.index === event.index);
+		const block = c.turnBlocks[index];
+		if (!block) return;
+		delete block.index;
+		if (block.type === "text") {
+			c.currentPiStream!.push({ type: "text_end", contentIndex: index, content: block.text, partial: c.turnOutput });
+		} else if (block.type === "thinking") {
+			c.currentPiStream!.push({ type: "thinking_end", contentIndex: index, content: block.thinking, partial: c.turnOutput });
+		} else if (block.type === "toolCall") {
+			c.turnSawToolCall = true;
+			block.arguments = mapToolArgs(
+				block.name, parsePartialJson(block.partialJson, block.arguments),
+			);
+			delete block.partialJson;
+			c.currentPiStream!.push({ type: "toolcall_end", contentIndex: index, toolCall: block, partial: c.turnOutput });
+		}
+		return;
+	}
+
+	if (event?.type === "message_delta") {
+		c.turnOutput.stopReason = mapStopReason(event.delta?.stop_reason);
+		if (event.usage) updateUsage(c.turnOutput, event.usage, model);
+		return;
+	}
+
+	if (event?.type === "message_stop" && c.turnSawToolCall) {
+		// Tool call complete — end this pi stream. The SDK will still yield an
+		// assistant message for this turn, but currentPiStream=null causes
+		// consumeQuery to skip it. The MCP handler blocks the generator until
+		// pi delivers the tool result via the next streamSimple call.
+		c.turnOutput.stopReason = "toolUse";
+		c.currentPiStream!.push({ type: "done", reason: "toolUse", message: c.turnOutput });
+		c.currentPiStream!.end();
+		c.currentPiStream = null;
+
+		// Cursor is updated by the next streamSimple call (tool result delivery path)
+		// which sets cursor = context.messages.length with the post-tool-result context.
+		return;
+	}
+
+	if (event?.type !== "message_stop" && event?.type !== "ping") {
+		debug("processStreamEvent: unhandled event type", event?.type);
+	}
+}
+
+// The SDK always yields `assistant` messages (completed content blocks) after streaming.
+// When stream_events already delivered the content, this is a no-op. But after
+// resetTurnState (e.g. tool result delivery), if the next turn's assistant message
+// arrives before any stream_events, this is the primary content path. Must maintain
+// the same stream lifecycle as processStreamEvent — including ending the stream on
+// tool_use to prevent deadlock with the MCP handler.
+function processAssistantMessage(message: SDKMessage, model: Model<any>, customToolNameToPi: Map<string, string>): void {
+	const c = ctx();
+	if (c.turnSawStreamEvent) return;
+	const assistantMsg = (message as any).message;
+	if (!assistantMsg?.content) return;
+	c.turnToolCallIds = [];
+	c.nextHandlerIdx = 0;
+	debug(`processAssistantMessage fallback: ${assistantMsg.content.length} blocks, types=${assistantMsg.content.map((b: any) => b.type).join(",")}`);
+	for (const block of assistantMsg.content) {
+		if (block.type === "text" && block.text) {
+			ensureTurnStarted();
+			c.turnBlocks.push({ type: "text", text: block.text });
+			const idx = c.turnBlocks.length - 1;
+			c.currentPiStream?.push({ type: "text_start", contentIndex: idx, partial: c.turnOutput });
+			c.currentPiStream?.push({ type: "text_delta", contentIndex: idx, delta: block.text, partial: c.turnOutput });
+			c.currentPiStream?.push({ type: "text_end", contentIndex: idx, content: block.text, partial: c.turnOutput });
+		} else if (block.type === "thinking") {
+			ensureTurnStarted();
+			c.turnBlocks.push({ type: "thinking", thinking: block.thinking ?? "", thinkingSignature: block.signature ?? "" });
+			const idx = c.turnBlocks.length - 1;
+			c.currentPiStream?.push({ type: "thinking_start", contentIndex: idx, partial: c.turnOutput });
+			if (block.thinking) c.currentPiStream?.push({ type: "thinking_delta", contentIndex: idx, delta: block.thinking, partial: c.turnOutput });
+			c.currentPiStream?.push({ type: "thinking_end", contentIndex: idx, content: block.thinking ?? "", partial: c.turnOutput });
+		} else if (block.type === "tool_use") {
+			ensureTurnStarted();
+			c.turnSawToolCall = true;
+			c.turnToolCallIds.push(block.id);
+			const mappedArgs = mapToolArgs(mapToolName(block.name, customToolNameToPi), block.input);
+			c.turnBlocks.push({
+				type: "toolCall", id: block.id,
+				name: mapToolName(block.name, customToolNameToPi),
+				arguments: mappedArgs,
+			});
+			const idx = c.turnBlocks.length - 1;
+			const toolBlock = c.turnBlocks[idx];
+			c.currentPiStream?.push({ type: "toolcall_start", contentIndex: idx, partial: c.turnOutput });
+			c.currentPiStream?.push({ type: "toolcall_end", contentIndex: idx, toolCall: toolBlock as any, partial: c.turnOutput });
+		} else {
+			debug("processAssistantMessage: unhandled block type", block.type);
+		}
+	}
+	if (assistantMsg.usage && c.turnOutput) updateUsage(c.turnOutput, assistantMsg.usage, model);
+
+	// End the stream on tool_use, same as processStreamEvent's message_stop handler.
+	if (c.turnSawToolCall && c.currentPiStream && c.turnOutput) {
+		c.turnOutput.stopReason = "toolUse";
+		c.currentPiStream.push({ type: "done", reason: "toolUse", message: c.turnOutput });
+		c.currentPiStream.end();
+		c.currentPiStream = null;
+	}
+}
+
+/** Background consumer: iterates the SDK generator, pushing events to currentPiStream.
+ *  Runs until the query ends. Per turn, the SDK yields stream_events (deltas), then
+ *  an assistant message (completed blocks). On tool_use, the stream is ended by
+ *  whichever path handles it first (processStreamEvent or processAssistantMessage),
+ *  and the MCP handler blocks the generator until pi delivers the tool result. */
+async function consumeQuery(
+	sdkQuery: ReturnType<typeof query>,
+	customToolNameToPi: Map<string, string>,
+	model: Model<any>,
+	wasAborted: () => boolean,
+): Promise<{ capturedSessionId?: string }> {
+	let capturedSessionId: string | undefined;
+
+	for await (const message of sdkQuery) {
+		if (wasAborted()) break;
+		if (!ctx().currentPiStream || !ctx().turnOutput) continue;
+
+		switch (message.type) {
+			case "stream_event":
+				processStreamEvent(message, customToolNameToPi, model);
+				break;
+			case "assistant":
+				processAssistantMessage(message, model, customToolNameToPi);
+				break;
+			case "result":
+				if (!ctx().turnSawStreamEvent && message.subtype === "success") {
+					ensureTurnStarted();
+					const text = message.result || "";
+					ctx().turnBlocks.push({ type: "text", text });
+					const idx = ctx().turnBlocks.length - 1;
+					ctx().currentPiStream?.push({ type: "text_start", contentIndex: idx, partial: ctx().turnOutput });
+					ctx().currentPiStream?.push({ type: "text_delta", contentIndex: idx, delta: text, partial: ctx().turnOutput });
+					ctx().currentPiStream?.push({ type: "text_end", contentIndex: idx, content: text, partial: ctx().turnOutput });
+				}
+				break;
+			case "system":
+				if ((message as any).subtype === "init" && (message as any).session_id) {
+					capturedSessionId = (message as any).session_id;
+				}
+				break;
+			case "user":
+				break; // SDK echo of user prompt — not needed
+			case "rate_limit_event": {
+				const info = (message as any).rate_limit_info;
+				debug("consumeQuery: rate_limit_event", JSON.stringify(info).slice(0, 300));
+				if (info?.status === "rejected") {
+					const resetsAt = info.resetsAt ? new Date(info.resetsAt).toLocaleTimeString() : "unknown";
+					piUI?.notify(`Claude rate limited (${info.rateLimitType ?? "unknown"}) — resets at ${resetsAt}`, "warning");
+				} else if (info?.status === "allowed_warning") {
+					piUI?.notify(`Claude rate limit warning: ${Math.round(info.utilization ?? 0)}% used (${info.rateLimitType ?? ""})`, "warning");
+				}
+				break;
+			}
+			default:
+				debug("consumeQuery: unhandled SDK message type", message.type);
+				break;
+		}
+	}
+
+	// DEBUG: trace when consumeQuery exits
+	debug(`consumeQuery: for-await loop exited, wasAborted=${wasAborted()}, capturedSessionId=${capturedSessionId?.slice(0, 8) ?? "none"}`);
+
+	return { capturedSessionId };
+}
+
+/** Provider entry point. Pi calls this for each new prompt and each tool result.
+ *  Two cases: tool result delivery (active query) or fresh query. */
+function streamClaudeAgentSdk(model: Model<any>, context: Context, options?: SimpleStreamOptions): AssistantMessageEventStream {
+	const stream = newAssistantMessageEventStream();
+
+	// DEBUG: trace followUp message triggering
+	const lastMsgRole = context.messages[context.messages.length - 1]?.role;
+	debug(`provider: streamClaudeAgentSdk called, activeQuery=${!!ctx().activeQuery}, lastMsgRole=${lastMsgRole}, isReentrant=${ctx().activeQuery !== null}`);
+
+	// --- Tool result delivery ---
+	// Pi appends tool results to context and calls back. Extract this turn's results
+	// (everything after the last assistant message) and match against waiting MCP
+	// handlers. Results that arrive before their handler get queued in pendingResults.
+	if (ctx().activeQuery) {
+		ctx().currentPiStream = stream;
+		ctx().resetTurnState(model);
+		const allResults = extractAllToolResults(context);
+		debug(`provider: tool results, ${allResults.length} results, ${ctx().pendingToolCalls.size} waiting handlers, ctx.msgs=${context.messages.length}`);
+		for (const result of allResults) {
+			const id = result.toolCallId;
+			if (id && ctx().pendingToolCalls.has(id)) {
+				const pending = ctx().pendingToolCalls.get(id)!;
+				ctx().pendingToolCalls.delete(id);
+				debug(`provider: resolving ${pending.toolName} [${id}]${result.isError ? " (error)" : ""}`, JSON.stringify(result.content).slice(0, 200));
+				pending.resolve(result);
+			} else if (id) {
+				ctx().pendingResults.set(id, result);
+				debug(`provider: queued result [${id}] (${ctx().pendingResults.size} pending)`);
+			} else {
+				debug(`WARNING: tool result without toolCallId, cannot match`);
+			}
+			if (ctx().pendingToolCalls.size > 0 && ctx().pendingResults.size > 0) {
+				debug(`BUG: both maps non-empty! handlers=${ctx().pendingToolCalls.size} results=${ctx().pendingResults.size}`);
+			}
+		}
+		if (ctx().pendingToolCalls.size > 0) {
+			debug(`WARNING: ${ctx().pendingToolCalls.size} MCP handlers still waiting after delivering ${allResults.length} results`);
+			piUI?.notify(`Claude bridge: ${ctx().pendingToolCalls.size} tool handler(s) still waiting — provider may be stuck`, "warning");
+		}
+
+		// Detect user messages (steer/followUp) that pi injected into context
+		// during the active query. This happens when:
+		//   - User sends a steer while a tool is executing; pi drains the steer
+		//     queue at the turn boundary and appends it to context alongside the
+		//     tool result, then calls the provider again.
+		//   - A followUp is delivered between tool-result turns.
+		// The bridge can't forward these mid-query (the SDK query is in progress),
+		// so we save them for replay as continuation queries after consumeQuery ends.
+		if (lastMsgRole === "user") {
+			const userPrompt = extractUserPrompt(context.messages);
+			if (userPrompt) {
+				ctx().deferredUserMessages.push(userPrompt);
+				debug(`provider: deferred user message for replay after query: ${userPrompt.slice(0, 60)}`);
+			}
+		}
+
+		if (sharedSession) sharedSession.cursor = context.messages.length;
+		ctx().latestCursor = Math.max(ctx().latestCursor, context.messages.length);
+		return stream;
+	}
+
+	// --- Orphaned tool result (e.g. user aborted a tool call) ---
+	// The query is gone but pi still delivered the result. Nothing to do — just
+	// emit end_turn so pi waits for the next real user message.
+	const lastMsg = context.messages[context.messages.length - 1];
+	if (lastMsg?.role === "toolResult") {
+		debug(`provider: orphaned tool result after abort, emitting end_turn`);
+		if (sharedSession) sharedSession.cursor = context.messages.length;
+		const c = ctx();  // capture current context for the microtask
+		queueMicrotask(() => {
+			c.resetTurnState(model);
+			stream.push({ type: "done", reason: "stop", message: c.turnOutput });
+			stream.end();
+		});
+		return stream;
+	}
+
+	// --- Fresh query ---
+
+	// 1. Determine reentrancy and push parent context if needed.
+	const isReentrant = ctx().activeQuery !== null;
+	if (isReentrant) pushContext();
+	debug(`provider: fresh query setup, isReentrant=${isReentrant}, stackDepth=${stackDepth()}`);
+
+	// 2. Fresh child context — constructor already gave us clean Maps and empty
+	//    arrays. For a reused top-level context, clear explicitly.
+	ctx().currentPiStream = stream;
+	ctx().pendingToolCalls.clear();
+	ctx().pendingResults.clear();
+	ctx().deferredUserMessages = [];
+	ctx().resetTurnState(model);
+	ctx().latestCursor = 0;
+
+	const { mcpTools, customToolNameToSdk, customToolNameToPi } = resolveMcpTools(context);
+	const cwd = (options as { cwd?: string } | undefined)?.cwd ?? process.cwd();
+	const { sessionId: resumeSessionId } = syncSharedSession(context.messages, cwd, customToolNameToSdk, model.id);
+	const promptBlocks = extractUserPromptBlocks(context.messages);
+	let promptText = extractUserPrompt(context.messages) ?? "";
+
+	// Guard: empty prompt means the last context message isn't a user message.
+	// This should never happen with the state stack fix — dump diagnostics if it does.
+	if (!promptText && !promptBlocks) {
+		diagDump("empty_prompt", {
+			contextLength: context.messages.length,
+			lastMsgRole: lastMsg?.role,
+			isReentrant,
+			stackDepth: stackDepth(),
+			activeQueryExists: ctx().activeQuery !== null,
+			sharedSession: sharedSession ? { sessionId: sharedSession.sessionId.slice(0, 8), cursor: sharedSession.cursor } : null,
+			messageRoles: context.messages.map((m, i) => `[${i}]${m.role}`).join(" "),
+		});
+		// Recover: use a continuation prompt so the SDK doesn't send an empty text block
+		promptText = "[continue]";
+	}
+
+	const prompt: string | AsyncIterable<SDKUserMessage> = promptBlocks
+		? wrapPromptStream(promptBlocks)
+		: promptText;
+	const mcpServers = buildMcpServers(mcpTools, ctx());
+	const bridgeConfig = loadConfig(cwd);
+	const providerSettings = bridgeConfig.provider ?? {};
+	const appendSystemPrompt = providerSettings.appendSystemPrompt !== false;
+	const agentsAppend = appendSystemPrompt ? extractAgentsAppend() : undefined;
+	const skillsAppend = appendSystemPrompt ? extractSkillsBlock(context.systemPrompt) : undefined;
+	const promptContextAppend = buildPromptContextAppend(context.systemPrompt, cwd, bridgeConfig.promptContext ?? {});
+	const appendParts = [agentsAppend, skillsAppend, promptContextAppend.text].filter((part): part is string => Boolean(part));
+	const systemPromptAppend = appendParts.length > 0 ? appendParts.join("\n\n") : undefined;
+
+	// MCP auto-loading suppression: with appendSystemPrompt=true (default), the
+	// SDK uses isolation mode and avoids filesystem settings. If users turn that
+	// off, load user/project settings but pass --strict-mcp-config so Claude Code
+	// ignores auto-discovered filesystem MCP servers while Pi owns tool execution.
+	const settingSources: SettingSource[] | undefined = appendSystemPrompt
+		? undefined
+		: providerSettings.settingSources ?? ["user", "project"];
+	const strictMcpConfigEnabled = !appendSystemPrompt && providerSettings.strictMcpConfig !== false;
+	const claudeExecutable = resolveClaudeExecutable(providerSettings.pathToClaudeCodeExecutable);
+
+	// Prefer the model's own thinkingLevelMap when present (pi-ai 0.72+ ships
+	// per-model overrides — e.g. opus-4-7 wants xhigh→xhigh, not xhigh→max).
+	// Fall back to our generic table for older pi-ai or unmapped levels.
+	const effort = options?.reasoning
+		? ((model as any).thinkingLevelMap?.[options.reasoning] as EffortLevel | undefined)
+			?? REASONING_TO_EFFORT[options.reasoning]
+		: undefined;
+
+	const extraArgs: Record<string, string | null> = { model: model.id };
+	if (strictMcpConfigEnabled) extraArgs["strict-mcp-config"] = null;
+	// Opus 4.7 defaults thinking.display to "omitted" (empty thinking text in stream).
+	// Force summarized so thinking_delta events arrive. See anthropics/claude-agent-sdk-python#830.
+	if (effort) extraArgs["thinking-display"] = "summarized";
+
+	// Suppress claude.ai cloud MCP servers (Figma/Canva/etc. auto-discovered via OAuth
+	// when the user is logged into Anthropic). These are a separate code path from
+	// filesystem MCP and are NOT blocked by --strict-mcp-config or settingSources=undefined.
+	// The native CC binary gates them on env var ENABLE_CLAUDEAI_MCP_SERVERS: setting it
+	// to "0"/"false"/"no"/"off" makes the loader return early before any cloud fetch.
+	// DISABLE_AUTO_COMPACT=1: pi owns context-management and propagates its own
+	// /compact via session_compact (see handler in default export). Letting CC
+	// also autocompact would double-flush the prompt cache and races pi's
+	// threshold with CC's, including CC's anti-thrashing guard (issue #8).
+	// Manual /compact in CC still works (we never invoke it).
+	const childEnv = { ...process.env, ENABLE_CLAUDEAI_MCP_SERVERS: "0", DISABLE_AUTO_COMPACT: "1" };
+	const queryOptions: NonNullable<Parameters<typeof query>[0]["options"]> = {
+		cwd,
+		env: childEnv,
+		disallowedTools: DISALLOWED_BUILTIN_TOOLS,
+		allowedTools: [`mcp__${MCP_SERVER_NAME}__*`],
+		permissionMode: "bypassPermissions",
+		includePartialMessages: true,
+		systemPrompt: {
+			type: "preset", preset: "claude_code",
+			append: systemPromptAppend ? systemPromptAppend : undefined,
+		},
+		extraArgs,
+		...(effort ? { effort } : {}),
+		...(settingSources ? { settingSources } : {}),
+		...(mcpServers ? { mcpServers } : {}),
+		...(resumeSessionId ? { resume: resumeSessionId } : {}),
+		...(claudeExecutable ? { pathToClaudeCodeExecutable: claudeExecutable } : {}),
+		...makeCliDebugOptions("provider"),
+	};
+
+	debug("provider: fresh query",
+		`model=${model.id} msgs=${context.messages.length} tools=${mcpTools.length}`,
+		`resume=${resumeSessionId?.slice(0, 8) ?? "none"} effort=${effort ?? "default"}`,
+		`appendSys=${appendSystemPrompt} promptCtx=${promptContextAppend.labels.join(",") || "none"} strictMcp=${strictMcpConfigEnabled}`,
+		`prompt=${promptText.slice(0, 60)}${promptBlocks ? " [+images]" : ""}`);
+
+	// 3. Start SDK query and claim it for this context
+	let wasAborted = false;
+	const sdkQuery = query({ prompt, options: queryOptions });
+	ctx().activeQuery = sdkQuery;
+
+	// 4. Capture context for abort handling (must be AFTER pushContext)
+	const abortCtx = ctx();
+
+	const requestAbort = () => {
+		// interrupt() asks the CLI to stop gracefully; close() kills it immediately.
+		// Both are needed — interrupt alone lets the current API call finish.
+		void sdkQuery.interrupt().catch(() => {});
+		try { sdkQuery.close(); } catch {}
+	};
+	const onAbort = () => {
+		wasAborted = true;
+		// Prevent stale deferred messages from being replayed by parent on pop
+		abortCtx.deferredUserMessages = [];
+		for (const pending of abortCtx.pendingToolCalls.values()) { pending.resolve({ content: [{ type: "text", text: "Operation aborted" }] }); }
+		abortCtx.pendingToolCalls.clear();
+		abortCtx.pendingResults.clear();
+		requestAbort();
+	};
+	if (options?.signal) {
+		if (options.signal.aborted) onAbort();
+		else options.signal.addEventListener("abort", onAbort, { once: true });
+	}
+
+	// Background consumer — runs until query ends
+	consumeQuery(sdkQuery, customToolNameToPi, model, () => wasAborted)
+		.then(async ({ capturedSessionId }) => {
+			debug(`provider: consumeQuery completed, stopReason=${ctx().turnOutput?.stopReason}, error=${ctx().turnOutput?.errorMessage}, aborted=${wasAborted}`);
+
+			// --- Abort detection in normal completion path ---
+			if (wasAborted || options?.signal?.aborted) {
+				if (sharedSession) sharedSession = { ...sharedSession, needsRebuild: true, forceRotate: true };
+				ctx().deferredUserMessages = [];
+				debug(`provider: abort detected, marked sharedSession needsRebuild + forceRotate`);
+				if (ctx().turnOutput) {
+					ctx().turnOutput.stopReason = "aborted";
+					ctx().turnOutput.errorMessage = "Operation aborted";
+				}
+				ctx().currentPiStream?.push({ type: "error", reason: "aborted", error: ctx().turnOutput! });
+				ctx().currentPiStream?.end();
+				ctx().currentPiStream = null;
+				return;
+			}
+
+			// --- Capture session ID ---
+			const sessionId = capturedSessionId ?? sharedSession?.sessionId;
+			if (sessionId) {
+				const cursor = Math.max(context.messages.length, ctx().latestCursor, sharedSession?.cursor ?? 0);
+				debug(`provider: query done, session=${sessionId.slice(0, 8)}, cursor=${cursor}`);
+				sharedSession = { sessionId, cursor, cwd };
+			}
+
+			// --- Replay deferred user messages as continuation queries ---
+			// Only for outermost queries — reentrant (subagent) queries leave
+			// deferred messages for the parent to handle after it finishes.
+			try {
+				while (ctx().deferredUserMessages.length > 0 && !isReentrant && !wasAborted) {
+					const steerPrompt = ctx().deferredUserMessages.shift()!;
+					debug(`provider: replaying deferred user message: ${steerPrompt.slice(0, 60)}`);
+					ctx().resetTurnState(model);
+
+					const resumeId = sharedSession?.sessionId;
+					if (!resumeId) {
+						debug(`WARNING: no session to resume for deferred message, dropping`);
+						break;
+					}
+
+					const contOptions = { ...queryOptions, resume: resumeId, ...makeCliDebugOptions("continuation") };
+					const contQuery = query({ prompt: steerPrompt, options: contOptions });
+					ctx().activeQuery = contQuery;
+
+					debug(`provider: continuation query, model=${model.id}, resume=${resumeId.slice(0, 8)}, prompt=${steerPrompt.slice(0, 60)}`);
+
+					try {
+						const { capturedSessionId: contSid } = await consumeQuery(contQuery, customToolNameToPi, model, () => wasAborted);
+						const sid = contSid ?? sharedSession?.sessionId;
+						if (sid) {
+							sharedSession = { sessionId: sid, cursor: sharedSession?.cursor ?? 0, cwd };
+						}
+					} catch (contError) {
+						debug(`provider: continuation query error:`, contError);
+						break;
+					} finally {
+						contQuery.close();
+					}
+				}
+			} finally {
+				// Guarantees restoration even if contQuery() throws synchronously
+				ctx().activeQuery = sdkQuery;
+			}
+
+			finalizeCurrentStream(ctx().turnOutput?.stopReason);
+		})
+		.catch((error) => {
+			debug(`provider: query error, model=${model.id}, aborted=${Boolean(options?.signal?.aborted)}, error=`, error);
+			if ((wasAborted || options?.signal?.aborted) && sharedSession) {
+				sharedSession = { ...sharedSession, needsRebuild: true, forceRotate: true };
+			} else {
+				sharedSession = null;
+			}
+			ctx().deferredUserMessages = [];
+			if (ctx().turnOutput) {
+				ctx().turnOutput.stopReason = options?.signal?.aborted ? "aborted" : "error";
+				ctx().turnOutput.errorMessage = error instanceof Error ? error.message : String(error);
+			}
+			ctx().currentPiStream?.push({ type: "error", reason: (ctx().turnOutput?.stopReason ?? "error") as "aborted" | "error", error: ctx().turnOutput! });
+			ctx().currentPiStream?.end();
+			ctx().currentPiStream = null;
+		})
+		.finally(() => {
+			if (options?.signal) options.signal.removeEventListener("abort", onAbort);
+			if (ctx().activeQuery === sdkQuery) {
+				// Drain pending handlers for this query
+				for (const pending of ctx().pendingToolCalls.values()) { pending.resolve({ content: [{ type: "text", text: "Query ended" }] }); }
+				ctx().pendingToolCalls.clear();
+				ctx().pendingResults.clear();
+
+				if (isReentrant) {
+					popContext();  // merges deferred messages and restores parent
+				} else {
+					ctx().activeQuery = null;
+				}
+			}
+			sdkQuery.close();
+		});
+
+	return stream;
+}
+
+// --- Extension registration ---
+
+export default function (pi: ExtensionAPI) {
+	// Disable non-essential Claude Code traffic (update checks, MCP registry, telemetry)
+	process.env.CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC = "1";
+
+	const config = loadConfig(process.cwd());
+	debug("loadConfig:", JSON.stringify(config));
+	if (config.enabled === false) {
+		debug("provider: disabled by configuration");
+		return;
+	}
+
+	// Reset shared session on pi session lifecycle events
+	const clearSession = (event: string) => {
+		debug(`${event}: clearing session ${sharedSession?.sessionId?.slice(0, 8) ?? "none"}`);
+		sharedSession = null;
+
+		// Clear the global streamSimple if this instance registered it.
+		// This allows /reload to work — the old instance clears the flag so
+		// the new instance can register fresh without wrapping stale state.
+		const g = globalThis as Record<symbol, any>;
+		if (g[ACTIVE_STREAM_SIMPLE_KEY] === streamClaudeAgentSdk) {
+			debug(`${event}: clearing ACTIVE_STREAM_SIMPLE_KEY`);
+			g[ACTIVE_STREAM_SIMPLE_KEY] = undefined;
+		}
+	};
+	pi.on("session_start", (event, ctx) => {
+		piUI = ctx.ui;
+		if (event.reason === "new" || event.reason === "resume" || event.reason === "fork") {
+			clearSession(`session_start:${event.reason}`);
+		}
+	});
+	pi.on("session_shutdown", () => clearSession("session_shutdown"));
+
+	// pi /compact and session-tree navigation (rewind / fork-at-point /
+	// branch switch) both mutate pi's messages array out from under the
+	// bridge. syncSharedSession's REUSE check would otherwise see
+	// slice(cursor) === [] (or skip entries) and keep --resume'ing a CC
+	// session that no longer matches pi's history. /compact in particular
+	// triggers CC's autocompact-thrashing guard (issue #8). Force the next
+	// call down the REBUILD path so CC sees the current history.
+	const markRebuild = (event: string) => {
+		if (sharedSession) {
+			debug(`${event}: marking needsRebuild on session ${sharedSession.sessionId.slice(0, 8)}`);
+			sharedSession = { ...sharedSession, needsRebuild: true };
+		}
+	};
+	pi.on("session_compact", () => markRebuild("session_compact"));
+	pi.on("session_tree", () => markRebuild("session_tree"));
+
+	// --- Provider ---
+	//
+	// Guard against re-registration when the module is loaded multiple times
+	// (e.g., when spawning subagents). The shared ModelRegistry would otherwise
+	// overwrite the parent's streamSimple, breaking tool result delivery.
+	// See ACTIVE_STREAM_SIMPLE_KEY for the full mechanism.
+
+	const g = globalThis as Record<symbol, any>;
+	if (!g[ACTIVE_STREAM_SIMPLE_KEY]) {
+		// First instance: store our streamSimple and register.
+		g[ACTIVE_STREAM_SIMPLE_KEY] = streamClaudeAgentSdk;
+		pi.registerProvider(PROVIDER_ID, {
+			baseUrl: "claude-bridge",
+			apiKey: "not-used",
+			api: "claude-bridge",
+			models: MODELS,
+			// Cast: pi-ai AssistantMessageEventStream diamond dep between pi-coding-agent and pi-agent-core
+			streamSimple: streamClaudeAgentSdk as any,
+		});
+	} else {
+		// Subsequent instance (subagent session): skip registration entirely.
+		// The subagent already has access to claude-bridge models via the shared
+		// ModelRegistry from the parent's registration. Calls to those models
+		// will route through the parent's streamSimple via the reentrant
+		// QueryContext stack mechanism.
+		debug(`provider: skipping re-registration, parent instance active (module=${moduleInstanceId})`);
+	}
+
+}