comisai 1.0.36 → 1.0.40

package/node_modules/@comis/daemon/dist/wiring/setup-agents/setup-agents-registry.d.ts

@@ -0,0 +1,141 @@
+/**
+ * Top-level agents-subsystem orchestrator: builds the `AgentsResult` registry
+ * by iterating over configured agents and delegating to setupSingleAgent.
+ * Constructs daemon-global state shared across every agent (OAuth credential
+ * store, provider health monitor, last-known-model tracker, file lock,
+ * encrypted OAuth profile store, periodic lock-cleanup timer).
+ *
+ * Imports `setupSingleAgent` + `SingleAgentDeps` from ./setup-agents-runtime.js
+ * and `resolveSubAgentToolNames` from ./setup-agents-tooling.js.
+ *
+ * @module
+ */
+import { type AppContainer, type InjectionRateLimiter, type OAuthCredentialStorePort, type SecretsCrypto, type ToolCapabilityPort } from "@comis/core";
+import type { ComisLogger } from "@comis/infra";
+import type Database from "better-sqlite3";
+import type { SqliteMemoryAdapter, createSessionStore } from "@comis/memory";
+import { createBudgetGuard, createCostTracker, createStepCounter, createSessionLifecycle, type AgentExecutor, type ActiveRunRegistry, type ProviderHealthMonitor } from "@comis/agent";
+import { type SkillRegistry, type SkillWatcherHandle, type McpClientManager } from "@comis/skills";
+import type { SingleAgentDeps } from "./setup-agents-types.js";
+import type { createComisSessionManager } from "@comis/agent";
+type PiSessionAdapter = ReturnType<typeof createComisSessionManager>;
+/** All services produced by the per-agent executor setup phase. */
+export interface AgentsResult {
+    /** Shared session manager across all agents. */
+    sessionManager: ReturnType<typeof createSessionLifecycle>;
+    /** Per-agent executor instances keyed by agentId. */
+    executors: Map<string, AgentExecutor>;
+    /** Per-agent workspace directory paths. */
+    workspaceDirs: Map<string, string>;
+    /** Per-agent cost trackers. */
+    costTrackers: Map<string, ReturnType<typeof createCostTracker>>;
+    /** Per-agent budget guards. */
+    budgetGuards: Map<string, ReturnType<typeof createBudgetGuard>>;
+    /** Per-agent step counters. */
+    stepCounters: Map<string, ReturnType<typeof createStepCounter>>;
+    /** Default agent ID from routing config. */
+    defaultAgentId: string;
+    /** Default agent workspace directory path. */
+    defaultWorkspaceDir: string;
+    /** Resolve executor for an agent ID, falling back to default agent. */
+    getExecutor: (agentId: string) => AgentExecutor;
+    /** Per-agent JSONL session adapters (keyed by agentId). */
+    piSessionAdapters: Map<string, PiSessionAdapter>;
+    /** Per-agent skill watcher handles for shutdown cleanup. */
+    skillWatcherHandles: Map<string, SkillWatcherHandle>;
+    /** Per-agent skill registries for skills.list RPC method. */
+    skillRegistries: Map<string, SkillRegistry>;
+    /**
+     * Per-agent ToolCapabilityPort instances. Consumed by setupTools via the
+     * getCapabilityPortForAgent closure on ToolsDeps; mutated by hot-add /
+     * hot-remove in daemon.ts to keep the parallel map consistent with
+     * skillRegistries.
+     */
+    toolCapabilityPorts: Map<string, ToolCapabilityPort>;
+    /** Periodic lock cleanup timer (cleared on shutdown). */
+    lockCleanupTimer: import("@comis/core").TimerHandle;
+    /** Shared single-agent dependencies for hot-add closure capture. */
+    singleAgentDeps: SingleAgentDeps;
+    /** Global provider health monitor for daemon-level health metrics */
+    providerHealth: ProviderHealthMonitor;
+    /**
+     * Daemon-level OAuthCredentialStore handle. Threaded into
+     * RpcDispatchDeps so agents.update can validate oauthProfiles patches
+     * via has().
+     */
+    oauthCredentialStore: OAuthCredentialStorePort;
+}
+/**
+ * Create the full per-agent executor subsystem: session manager, per-agent
+ * workspace directories, safety dependencies (circuit breaker, budget guard,
+ * cost tracker, step counter), and PiExecutor instances.
+ * @param deps.container      - Bootstrap output (config, event bus, secret manager)
+ * @param deps.memoryAdapter  - SQLite memory adapter (from setupMemory result)
+ * @param deps.sessionStore   - Session persistence store (from setupMemory result)
+ * @param deps.agentLogger    - Module-bound logger for agent subsystem
+ */
+export declare function setupAgents(deps: {
+    container: AppContainer;
+    memoryAdapter: SqliteMemoryAdapter;
+    sessionStore: ReturnType<typeof createSessionStore>;
+    agentLogger: ComisLogger;
+    /** When true, executor includes MEDIA: directive instructions in system prompt. */
+    outboundMediaEnabled?: boolean;
+    /** When true, executor system prompt includes attachment hint processing guidance.
+     * Set to true when at least one auto-processing pipeline is disabled. */
+    autonomousMediaEnabled?: boolean;
+    /** Optional active run registry for SDK-native steer+followup */
+    activeRunRegistry?: ActiveRunRegistry;
+    /** Fallback secret for canary token generation when CANARY_SECRET not configured. */
+    canaryFallbackSecret?: string;
+    /** Injection rate limiter singleton for progressive cooldown (optional). */
+    injectionRateLimiter?: InjectionRateLimiter;
+    /** Embedding queue for async vector generation. Wired into executor for conversation persistence. */
+    embeddingQueue?: {
+        enqueue(entryId: string, content: string): void;
+    };
+    /** Context store for DAG mode context engine */
+    contextStore?: import("@comis/core").ContextStorePort;
+    /** Raw better-sqlite3 database handle for DAG transactions */
+    db?: unknown;
+    /** Optional embedding port for discover_tools semantic search. */
+    embeddingPort?: import("@comis/core").EmbeddingPort;
+    /** Delivery mirror port for session mirroring injection */
+    deliveryMirror?: import("@comis/core").DeliveryMirrorPort;
+    /** Delivery mirror config for injection budget */
+    deliveryMirrorConfig?: {
+        maxEntriesPerInjection: number;
+        maxCharsPerInjection: number;
+    };
+    /** Gemini CachedContent lifecycle manager. */
+    geminiCacheManager?: import("@comis/agent").GeminiCacheManager;
+    /** Resolve platform message character limit for a channel type. */
+    getChannelMaxChars?: (channelType: string) => number | undefined;
+    /** Background task manager for auto-promotion of long-running tools. */
+    backgroundTaskManager?: import("@comis/agent").BackgroundTaskManager;
+    /**
+     * SecretsCrypto engine bound to SECRETS_MASTER_KEY. Defined when daemon
+     * was started with a valid master key. Required for
+     * `appConfig.oauth.storage === "encrypted"` mode.
+     */
+    secretsCrypto?: SecretsCrypto;
+    /**
+     * Shared better-sqlite3 handle to secrets.db. Plumbed from daemon.ts where
+     * createSqliteSecretStore now exposes its db field. Required for
+     * `appConfig.oauth.storage === "encrypted"` mode.
+     */
+    secretsDb?: Database.Database;
+    /**
+     * Daemon-global MCP client manager. setupSingleAgent constructs a
+     * per-agent ToolCapabilityPort adapter that closes over this manager.
+     * daemon.ts threads it in after running setupMcp before setupAgents.
+     */
+    mcpClientManager: McpClientManager;
+    /** Wall-clock + monotonic time reads. */
+    clock: import("@comis/core").ClockPort;
+    /** Environment-variable reads. */
+    env: import("@comis/core").EnvPort;
+    /** Timer scheduling. */
+    timers: import("@comis/core").TimerPort;
+}): Promise<AgentsResult>;
+export {};

package/node_modules/@comis/daemon/dist/wiring/setup-agents/setup-agents-registry.js

@@ -0,0 +1,278 @@
+// SPDX-License-Identifier: Apache-2.0
+// @allow-throw: setup-agents registry guards; consumed at daemon.ts bootstrap catch boundary.
+/**
+ * Top-level agents-subsystem orchestrator: builds the `AgentsResult` registry
+ * by iterating over configured agents and delegating to setupSingleAgent.
+ * Constructs daemon-global state shared across every agent (OAuth credential
+ * store, provider health monitor, last-known-model tracker, file lock,
+ * encrypted OAuth profile store, periodic lock-cleanup timer).
+ *
+ * Imports `setupSingleAgent` + `SingleAgentDeps` from ./setup-agents-runtime.js
+ * and `resolveSubAgentToolNames` from ./setup-agents-tooling.js.
+ *
+ * @module
+ */
+import { safePath } from "@comis/core";
+// The encrypted-store factory is value-imported here because daemon owns
+// secretsDb + secretsCrypto. Constructing the store at this composition
+// site keeps agent free of any production @comis/memory import.
+import { createOAuthProfileStoreEncrypted } from "@comis/memory";
+import { homedir } from "node:os";
+import { existsSync, mkdirSync } from "node:fs";
+import { suppressError } from "@comis/shared";
+import { createSessionLifecycle, cleanupStaleLocks, createProviderHealthMonitor, createLastKnownModelTracker, setSanitizeLogger, setToolNormalizationLogger, } from "@comis/agent";
+// Symbols imported directly from @comis/core — the daemon composition
+// root no longer goes through @comis/agent re-exports.
+import { selectOAuthCredentialStore, 
+// Canonical FileLockPort adapter consumed here as the production
+// createFileLock() target so the daemon no longer reaches into
+// @comis/scheduler for it.
+createFileLock, } from "@comis/core";
+import { setupSingleAgent } from "./setup-agents-runtime.js";
+import { resolveSubAgentToolNames } from "./setup-agents-tooling.js";
+// Once-per-daemon-process WARN flag for the encrypted-store hot-reload
+// limitation. Lifted to module scope so the flag survives across per-agent
+// setupSingleAgent calls AND any future re-invocations of setupAgents within
+// the same process. Operator-friendly notice — fires exactly once per daemon
+// process so the operator sees it in startup logs without N-times-per-agent
+// noise.
+let encryptedModeWarnFired = false;
+// ---------------------------------------------------------------------------
+// Setup function
+// ---------------------------------------------------------------------------
+/**
+ * Create the full per-agent executor subsystem: session manager, per-agent
+ * workspace directories, safety dependencies (circuit breaker, budget guard,
+ * cost tracker, step counter), and PiExecutor instances.
+ * @param deps.container      - Bootstrap output (config, event bus, secret manager)
+ * @param deps.memoryAdapter  - SQLite memory adapter (from setupMemory result)
+ * @param deps.sessionStore   - Session persistence store (from setupMemory result)
+ * @param deps.agentLogger    - Module-bound logger for agent subsystem
+ */
+export async function setupAgents(deps) {
+    const { container, memoryAdapter, sessionStore, agentLogger } = deps;
+    // Inject module-level logger for response sanitization pipeline
+    setSanitizeLogger(agentLogger.child({ submodule: "response-sanitize" }));
+    // Inject module-level logger for tool schema normalization pipeline
+    setToolNormalizationLogger(agentLogger.child({ submodule: "tool-normalize" }));
+    // Once-per-daemon WARN for the encrypted-store hot-reload limitation.
+    // Placed in setupAgents() body (NOT setupSingleAgent) so the notice fires
+    // exactly once per daemon process — not N times for N agents. Operator
+    // sees this in startup logs without surprise; daemon restart is required
+    // to pick up CLI-written OAuth profiles in encrypted-store mode.
+    const overallStorageMode = container.config.oauth.storage;
+    if (overallStorageMode === "encrypted" && !encryptedModeWarnFired) {
+        encryptedModeWarnFired = true;
+        agentLogger.warn({
+            hint: "CLI auth login changes require daemon restart in encrypted mode (file-watch unsupported on encrypted SQLite WAL)",
+            errorKind: "config",
+            submodule: "setup-agents",
+        }, "OAuth hot-reload disabled in encrypted-store mode");
+    }
+    const agents = container.config.agents; // Always populated after schema transform
+    const routingConfig = container.config.routing;
+    // Daemon-level tracing defaults
+    const daemonTracingDefaults = container.config.daemon?.logging?.tracing;
+    // Resolve agentDir for SDK persistent settings (root config, defaults to ~/.pi/agent)
+    const agentDir = container.config.agentDir;
+    const resolvedAgentDir = agentDir.startsWith("~")
+        ? agentDir.replace("~", homedir())
+        : agentDir;
+    agentLogger.debug({ agentDir: resolvedAgentDir }, "SDK agent directory resolved");
+    // Auto-create agentDir if missing (SDK needs this directory for settings files)
+    try {
+        if (!existsSync(resolvedAgentDir)) {
+            mkdirSync(resolvedAgentDir, { recursive: true });
+            agentLogger.info({ agentDir: resolvedAgentDir }, "Created SDK agent directory");
+        }
+    }
+    catch (mkdirError) {
+        agentLogger.warn({
+            agentDir: resolvedAgentDir,
+            err: mkdirError,
+            hint: "Failed to create agentDir; SettingsManager will fall back to in-memory",
+            errorKind: "config",
+        }, "Agent directory creation failed");
+    }
+    // Create shared services (session manager is shared across agents)
+    const sessionManager = createSessionLifecycle(sessionStore);
+    // Per-agent executor map
+    const executors = new Map();
+    const workspaceDirs = new Map();
+    const costTrackers = new Map();
+    const budgetGuards = new Map();
+    const stepCounters = new Map();
+    const piSessionAdapters = new Map();
+    const skillWatcherHandles = new Map();
+    const skillRegistries = new Map();
+    // Per-agent live ToolCapabilityPort adapters, parallel to skillRegistries;
+    // mutated in lockstep by hot-add/hot-remove.
+    const toolCapabilityPorts = new Map();
+    // Resolve sub-agent tool names from config for delegation awareness
+    const subAgentToolGroups = container.config.security?.agentToAgent?.subAgentToolGroups ?? [];
+    const subAgentToolNames = subAgentToolGroups.length === 0 || subAgentToolGroups.includes("full")
+        ? undefined // Full profile or unconfigured = all tools, no need for awareness section
+        : resolveSubAgentToolNames(subAgentToolGroups);
+    // MCP-AWARE: Whether sub-agents inherit MCP tools (used in system prompt to avoid false "do NOT have" claims)
+    const mcpToolsInherited = (container.config.security?.agentToAgent?.subAgentMcpTools ?? "inherit") === "inherit";
+    // Global provider health monitor (shared across all agents)
+    const providerHealth = createProviderHealthMonitor({
+        degradedThreshold: 2,
+        consecutiveFailureThreshold: 3,
+        windowMs: 60_000,
+        recoveryThreshold: 1,
+        eventBus: container.eventBus,
+    });
+    // Global last-known-working model tracker (shared across all agents)
+    const lastKnownModel = createLastKnownModelTracker();
+    // Construct the daemon-level OAuthCredentialStore handle ONCE (instead of
+    // per-agent inside setupSingleAgent). Same handle is reused across every
+    // agent setup AND surfaced on AgentsResult so daemon.ts can plumb it into
+    // RpcDispatchDeps for the agents.update oauthProfiles existence check.
+    const dataDirAbsForOauth = container.config.dataDir && container.config.dataDir.length > 0
+        ? container.config.dataDir
+        : safePath(homedir(), ".comis");
+    // Construct the encrypted-mode store HERE — daemon already owns secretsDb
+    // + secretsCrypto, so the agent selector does not reach into @comis/memory.
+    //
+    // EXPLICIT GUARDS (NOT non-null assertions): if storage is "encrypted" but
+    // either secretsDb or secretsCrypto is unset, throw a clear bootstrap error
+    // pointing operators at SECRETS_MASTER_KEY. Non-null assertions (!) would
+    // produce a `TypeError: Cannot read properties of undefined` at runtime —
+    // exactly the failure mode the old in-selector pre-check avoided.
+    let encryptedStore;
+    if (container.config.oauth.storage === "encrypted") {
+        if (!deps.secretsDb || !deps.secretsCrypto) {
+            throw new Error("OAuth storage mode is 'encrypted' but secretsDb/secretsCrypto were not initialized. " +
+                "Hint: set SECRETS_MASTER_KEY env var (and restart the daemon) so the encrypted " +
+                "secrets store boots, or change appConfig.oauth.storage to 'file' to use the " +
+                "plaintext file backend.");
+        }
+        encryptedStore = createOAuthProfileStoreEncrypted(deps.secretsDb, deps.secretsCrypto);
+    }
+    // Construct the canonical FileLockPort adapter ONCE here. Reused for
+    // OAuth store/manager locking AND session-write-lock + stale-lock
+    // cleanup across every per-agent setup. The port is stateless (per
+    // `createFileLock` semantics in @comis/scheduler), so a
+    // single shared instance is correct.
+    const fileLock = createFileLock();
+    const oauthCredentialStore = selectOAuthCredentialStore({
+        storage: container.config.oauth.storage,
+        dataDir: dataDirAbsForOauth,
+        fileLock,
+        encryptedStore,
+    });
+    // Construct shared deps struct once before the loop (for hot-add reuse)
+    const singleAgentDeps = {
+        container,
+        memoryAdapter,
+        sessionStore,
+        agentLogger,
+        resolvedAgentDir,
+        daemonTracingDefaults,
+        subAgentToolNames,
+        mcpToolsInherited,
+        outboundMediaEnabled: deps.outboundMediaEnabled,
+        autonomousMediaEnabled: deps.autonomousMediaEnabled,
+        activeRunRegistry: deps.activeRunRegistry,
+        canaryFallbackSecret: deps.canaryFallbackSecret,
+        injectionRateLimiter: deps.injectionRateLimiter,
+        embeddingQueue: deps.embeddingQueue,
+        contextStore: deps.contextStore,
+        db: deps.db,
+        providerHealth,
+        lastKnownModel,
+        embeddingPort: deps.embeddingPort,
+        deliveryMirror: deps.deliveryMirror,
+        deliveryMirrorConfig: deps.deliveryMirrorConfig,
+        geminiCacheManager: deps.geminiCacheManager,
+        getChannelMaxChars: deps.getChannelMaxChars,
+        backgroundTaskManager: deps.backgroundTaskManager,
+        // Secrets bootstrap output for OAuth wiring.
+        secretsCrypto: deps.secretsCrypto,
+        secretsDb: deps.secretsDb,
+        // Daemon-level OAuth credential store handle (constructed once above,
+        // reused per-agent + threaded into RPC deps).
+        oauthCredentialStore,
+        // Daemon-global MCP client manager. Threaded through so each
+        // setupSingleAgent invocation can construct a per-agent
+        // ToolCapabilityPort adapter that closes over the live MCP state.
+        mcpClientManager: deps.mcpClientManager,
+        // Canonical FileLockPort adapter — see comment at construction site.
+        fileLock,
+        // Runtime adapter ports.
+        clock: deps.clock,
+        env: deps.env,
+        timers: deps.timers,
+    };
+    for (const [agentId, agentConfig] of Object.entries(agents)) {
+        const result = await setupSingleAgent(agentId, agentConfig, singleAgentDeps);
+        executors.set(agentId, result.executor);
+        workspaceDirs.set(agentId, result.workspaceDir);
+        costTrackers.set(agentId, result.costTracker);
+        budgetGuards.set(agentId, result.budgetGuard);
+        stepCounters.set(agentId, result.stepCounter);
+        piSessionAdapters.set(agentId, result.piSessionAdapter);
+        if (result.skillWatcherHandle)
+            skillWatcherHandles.set(agentId, result.skillWatcherHandle);
+        skillRegistries.set(agentId, result.skillRegistry);
+        toolCapabilityPorts.set(agentId, result.toolCapabilityPort);
+    }
+    const defaultAgentId = routingConfig.defaultAgentId;
+    const defaultWorkspaceDir = workspaceDirs.get(defaultAgentId);
+    /** Resolve executor for an agent ID, falling back to default agent. */
+    function getExecutor(agentId) {
+        const exec = executors.get(agentId);
+        if (!exec) {
+            const fallback = executors.get(defaultAgentId);
+            if (!fallback)
+                throw new Error(`No executor found for agent: ${agentId}`);
+            return fallback;
+        }
+        return exec;
+    }
+    // Periodic stale lock cleanup (every 30 minutes)
+    const LOCK_CLEANUP_INTERVAL_MS = 30 * 60_000;
+    const lockCleanupTimer = deps.timers.setInterval(() => {
+        for (const [agentId, dir] of workspaceDirs) {
+            const lockDir = safePath(dir, ".locks");
+            suppressError(cleanupStaleLocks(fileLock, lockDir).then((removed) => {
+                if (removed > 0) {
+                    agentLogger.info({ agentId, removed, lockDir }, "Periodic stale lock cleanup");
+                }
+            }), "periodic lock cleanup");
+        }
+    }, LOCK_CLEANUP_INTERVAL_MS);
+    // Prevent timer from keeping the process alive during shutdown
+    lockCleanupTimer.unref();
+    return {
+        sessionManager,
+        executors,
+        workspaceDirs,
+        costTrackers,
+        budgetGuards,
+        stepCounters,
+        defaultAgentId,
+        defaultWorkspaceDir,
+        getExecutor,
+        piSessionAdapters,
+        skillWatcherHandles,
+        skillRegistries,
+        // daemon.ts threads this Map into setupTools via
+        // getCapabilityPortForAgent and mutates it on hot-add/hot-remove.
+        toolCapabilityPorts,
+        lockCleanupTimer,
+        singleAgentDeps,
+        providerHealth,
+        // Daemon-level OAuth credential store, plumbed by daemon.ts into
+        // RpcDispatchDeps.oauthCredentialStore so agents.update can validate
+        // oauthProfiles patches via has().
+        oauthCredentialStore,
+    };
+}
+// ---------------------------------------------------------------------------
+// OAuth credential store selection lives in @comis/agent (CLI cannot import
+// from @comis/daemon, so the helper must live where both daemon and CLI
+// consume it).
+// See: packages/agent/src/model/oauth-credential-store-selector.ts
+// ---------------------------------------------------------------------------

package/node_modules/@comis/daemon/dist/wiring/setup-agents/setup-agents-runtime.d.ts

@@ -0,0 +1,31 @@
+/**
+ * Per-agent executor runtime: session manager, per-agent workspace, safety
+ * dependencies (circuit breaker, budget guard, cost tracker, step counter),
+ * and PiExecutor creation. Single-agent factory `setupSingleAgent` and the
+ * shared `SingleAgentDeps` / `SingleAgentResult` interface declarations.
+ *
+ * Imports `resolveAgentModel` + `deriveCanaryFallback` from
+ * ./setup-agents-tooling.js; the top-level orchestration loop lives in
+ * ./setup-agents-registry.js.
+ *
+ * The live ToolCapabilityPort adapter is constructed inside setupSingleAgent
+ * (this function), NOT at a higher composition site (daemon.ts). Rationale:
+ * skillRegistry is per-agent and the skill-allow/deny precedence chain is
+ * per-agent; a daemon-global adapter cannot satisfy this without breaking
+ * the port interface (would require adding agentId to every method). The
+ * per-agent port is exposed via AgentsResult.toolCapabilityPorts and threaded
+ * into setupTools via the getCapabilityPortForAgent closure on ToolsDeps.
+ *
+ * @module
+ */
+import { type PerAgentConfig } from "@comis/core";
+import type { SingleAgentDeps, SingleAgentResult } from "./setup-agents-types.js";
+export type { SingleAgentDeps, SingleAgentResult } from "./setup-agents-types.js";
+/**
+ * Set up a single agent's executor and all supporting services.
+ * Validates rawAgentConfig with PerAgentConfigSchema before any runtime setup.
+ * On validation failure the Zod error propagates to the caller.
+ * Extracted from the setupAgents() loop body so it can be called independently
+ * for hot-add (adding an agent at runtime without daemon restart).
+ */
+export declare function setupSingleAgent(agentId: string, rawAgentConfig: PerAgentConfig, deps: SingleAgentDeps): Promise<SingleAgentResult>;