TaskFlow
DashboardFreewriteWhiteboardsProjectsCRMTasksNotificationsSettingsAgent TowerAPI Docs
OpenClaw Docs
?

User

Member

Caricamento in corso...

Home
Progetti
Task
Notifiche
CRM

    OpenClaw

    Documentation Mirror

    Documentation Overview

    Docs

    Auth credential semantics
    Scheduled tasks
    Hooks
    Automation & tasks
    Standing orders
    Task flow
    Background tasks
    BlueBubbles
    Broadcast groups
    Channel routing
    Discord
    Feishu
    Google Chat
    Group messages
    Groups
    iMessage
    Chat channels
    IRC
    LINE
    Channel location parsing
    Matrix
    Matrix migration
    Matrix push rules for quiet previews
    Mattermost
    Microsoft Teams
    Nextcloud Talk
    Nostr
    Pairing
    QA channel
    QQ bot
    Signal
    Slack
    Synology Chat
    Telegram
    Tlon
    Channel troubleshooting
    Twitch
    WeChat
    WhatsApp
    Yuanbao
    Zalo
    Zalo personal
    CI pipeline
    ACP
    Agent
    Agents
    Approvals
    Backup
    Browser
    Channels
    Clawbot
    `openclaw commitments`
    Completion
    Config
    Configure
    Cron
    Daemon
    Dashboard
    Devices
    Directory
    DNS
    Docs
    Doctor
    Flows (redirect)
    Gateway
    Health
    Hooks
    CLI reference
    Inference CLI
    Logs
    MCP
    Memory
    Message
    Migrate
    Models
    Node
    Nodes
    Onboard
    Pairing
    Plugins
    Proxy
    QR
    Reset
    Sandbox CLI
    Secrets
    Security
    Sessions
    Setup
    Skills
    Status
    System
    `openclaw tasks`
    TUI
    Uninstall
    Update
    Voicecall
    Webhooks
    Wiki
    Active memory
    Agent runtime
    Agent loop
    Agent runtimes
    Agent workspace
    Gateway architecture
    Channel docking
    Inferred commitments
    Compaction
    Context
    Context engine
    Delegate architecture
    Dreaming
    Experimental features
    Features
    Markdown formatting
    Memory overview
    Builtin memory engine
    Honcho memory
    QMD memory engine
    Memory search
    Messages
    Model failover
    Model providers
    Models CLI
    Multi-agent routing
    OAuth
    OpenClaw App SDK
    Presence
    QA overview
    Matrix QA
    Command queue
    Steering queue
    Retry policy
    Session management
    Session pruning
    Session tools
    SOUL.md personality guide
    Streaming and chunking
    System prompt
    Timezones
    TypeBox
    Typing indicators
    Usage tracking
    Date and time
    Node + tsx crash
    Diagnostics flags
    Authentication
    Background exec and process tool
    Bonjour discovery
    Bridge protocol
    CLI backends
    Configuration — agents
    Configuration — channels
    Configuration — tools and custom providers
    Configuration
    Configuration examples
    Configuration reference
    Diagnostics export
    Discovery and transports
    Doctor
    Gateway lock
    Health checks
    Heartbeat
    Gateway runbook
    Local models
    Gateway logging
    Multiple gateways
    Network model
    OpenAI chat completions
    OpenResponses API
    OpenShell
    OpenTelemetry export
    Gateway-owned pairing
    Prometheus metrics
    Gateway protocol
    Remote access
    Remote gateway setup
    Sandbox vs tool policy vs elevated
    Sandboxing
    Secrets management
    Secrets apply plan contract
    Security audit checks
    Security
    Tailscale
    Tools invoke API
    Troubleshooting
    Trusted proxy auth
    Debugging
    Environment variables
    FAQ
    FAQ: first-run setup
    FAQ: models and auth
    GPT-5.5 / Codex agentic parity
    GPT-5.5 / Codex parity maintainer notes
    Help
    Scripts
    Testing
    Testing: live suites
    General troubleshooting
    OpenClaw
    Ansible
    Azure
    Bun (experimental)
    ClawDock
    Release channels
    DigitalOcean
    Docker
    Docker VM runtime
    exe.dev
    Fly.io
    GCP
    Hetzner
    Hostinger
    Install
    Installer internals
    Kubernetes
    macOS VMs
    Migration guide
    Migrating from Claude
    Migrating from Hermes
    Nix
    Node.js
    Northflank
    Oracle Cloud
    Podman
    Railway
    Raspberry Pi
    Render
    Uninstall
    Updating
    Logging
    Network
    Audio and voice notes
    Camera capture
    Image and media support
    Nodes
    Location command
    Media understanding
    Talk mode
    Node troubleshooting
    Voice wake
    Pi integration architecture
    Pi development workflow
    Android app
    Platforms
    iOS app
    Linux app
    Gateway on macOS
    Canvas
    Gateway lifecycle
    macOS dev setup
    Health checks (macOS)
    Menu bar icon
    macOS logging
    Menu bar
    Peekaboo bridge
    macOS permissions
    Remote control
    macOS signing
    Skills (macOS)
    Voice overlay
    Voice wake (macOS)
    WebChat (macOS)
    macOS IPC
    macOS app
    Windows
    Plugin internals
    Plugin architecture internals
    Building plugins
    Plugin bundles
    Codex Computer Use
    Codex harness
    Community plugins
    Plugin compatibility
    Plugin dependency resolution
    Google Meet plugin
    Plugin hooks
    Plugin manifest
    Memory LanceDB
    Memory wiki
    Message presentation
    Agent harness plugins
    Building channel plugins
    Channel turn kernel
    Plugin entry points
    Plugin SDK migration
    Plugin SDK overview
    Building provider plugins
    Plugin runtime helpers
    Plugin setup and config
    Plugin SDK subpaths
    Plugin testing
    Skill workshop plugin
    Voice call plugin
    Webhooks plugin
    Zalo personal plugin
    OpenProse
    Alibaba Model Studio
    Anthropic
    Arcee AI
    Azure Speech
    Amazon Bedrock
    Amazon Bedrock Mantle
    Chutes
    Claude Max API proxy
    Cloudflare AI gateway
    ComfyUI
    Deepgram
    Deepinfra
    DeepSeek
    ElevenLabs
    Fal
    Fireworks
    GitHub Copilot
    GLM (Zhipu)
    Google (Gemini)
    Gradium
    Groq
    Hugging Face (inference)
    Provider directory
    Inferrs
    Inworld
    Kilocode
    LiteLLM
    LM Studio
    MiniMax
    Mistral
    Model provider quickstart
    Moonshot AI
    NVIDIA
    Ollama
    OpenAI
    OpenCode
    OpenCode Go
    OpenRouter
    Perplexity
    Qianfan
    Qwen
    Runway
    SGLang
    StepFun
    Synthetic
    Tencent Cloud (TokenHub)
    Together AI
    Venice AI
    Vercel AI gateway
    vLLM
    Volcengine (Doubao)
    Vydra
    xAI
    Xiaomi MiMo
    Z.AI
    Default AGENTS.md
    Release policy
    API usage and costs
    Credits
    Device model database
    Full release validation
    Memory configuration reference
    OpenClaw App SDK API design
    Prompt caching
    Rich output protocol
    RPC adapters
    SecretRef credential surface
    Session management deep dive
    AGENTS.md template
    BOOT.md template
    BOOTSTRAP.md template
    HEARTBEAT.md template
    IDENTITY template
    SOUL.md template
    TOOLS.md template
    USER template
    Tests
    Token use and costs
    Transcript hygiene
    Onboarding reference
    Contributing to the threat model
    Threat model (MITRE ATLAS)
    Formal verification (security models)
    Network proxy
    Agent bootstrapping
    Docs directory
    Getting started
    Docs hubs
    OpenClaw lore
    Onboarding (macOS app)
    Onboarding overview
    Personal assistant setup
    Setup
    Showcase
    Onboarding (CLI)
    CLI automation
    CLI setup reference
    ACP agents
    ACP agents — setup
    Agent send
    apply_patch tool
    Brave search
    Browser (OpenClaw-managed)
    Browser control API
    Browser troubleshooting
    Browser login
    WSL2 + Windows + remote Chrome CDP troubleshooting
    BTW side questions
    ClawHub
    Code execution
    Creating skills
    Diffs
    DuckDuckGo search
    Elevated mode
    Exa search
    Exec tool
    Exec approvals
    Exec approvals — advanced
    Firecrawl
    Gemini search
    Grok search
    Image generation
    Tools and plugins
    Kimi search
    LLM task
    Lobster
    Tool-loop detection
    Media overview
    MiniMax search
    Multi-agent sandbox and tools
    Music generation
    Ollama web search
    PDF tool
    Perplexity search
    Plugins
    Reactions
    SearXNG search
    Skills
    Skills config
    Slash commands
    Sub-agents
    Tavily
    Thinking levels
    Tokenjuice
    Trajectory bundles
    Text-to-speech
    Video generation
    Web search
    Web fetch
    Linux server
    Control UI
    Dashboard
    Web
    TUI
    WebChat

    OpenAPI Specs

    openapi
    TaskFlow
    docs/openclaw
    Original Docs

    Real-time Synchronized Documentation

    Last sync: 01/05/2026 09:36:57

    Note: This content is mirrored from docs.openclaw.ai and is subject to their terms and conditions.

    OpenClaw Docs

    v2.4.0 Production

    Last synced: Today, 22:00

    Technical reference for the OpenClaw framework. Real-time synchronization with the official documentation engine.

    Use this file to discover all available pages before exploring further.

    Plugin dependency resolution

    OpenClaw does not install every bundled plugin dependency tree at package install time. It first derives an effective plugin plan from config and plugin metadata, then stages runtime dependencies only for bundled OpenClaw-owned plugins that the plan can actually load.

    This page covers packaged runtime dependencies for bundled OpenClaw plugins. Third-party plugins and custom plugin paths still use explicit plugin installation commands such as

    text
    openclaw plugins install
    and
    text
    openclaw plugins update
    .

    Responsibility split

    OpenClaw owns the plan and policy:

    • which plugins are active for this config
    • which dependency roots are writable or read-only
    • when repair is allowed
    • which plugin ids are staged for startup
    • final checks before importing plugin runtime modules

    The package manager owns dependency convergence:

    • package graph resolution
    • production, optional, and peer dependency handling
    • text
      node_modules
      layout
    • package integrity
    • lock and install metadata

    In practice, OpenClaw should decide what needs to exist.

    text
    pnpm
    or
    text
    npm
    should make the filesystem match that decision.

    OpenClaw also owns the per-install-root coordination lock. Package managers protect their own install transaction, but they do not serialize OpenClaw's manifest writes, isolated-stage copy/rename, final validation, or plugin import against another Gateway, doctor, or CLI process touching the same runtime dependency root.

    Effective plugin plan

    The effective plugin plan is derived from config plus discovered plugin metadata. These inputs can activate bundled plugin runtime dependencies:

    • text
      plugins.entries.<id>.enabled
    • text
      plugins.allow
      ,
      text
      plugins.deny
      , and
      text
      plugins.enabled
    • legacy channel config such as
      text
      channels.telegram.enabled
    • configured providers, models, or CLI backend references that require a plugin
    • bundled manifest defaults such as
      text
      enabledByDefault
    • the installed plugin index and bundled manifest metadata

    Explicit disablement wins. A disabled plugin, denied plugin id, disabled plugin system, or disabled channel does not trigger runtime dependency repair. Persisted auth state alone also does not activate a bundled channel or provider.

    The plugin plan is the stable input. The generated dependency materialization is an output of that plan.

    Startup flow

    Gateway startup parses config and builds the startup plugin lookup table before plugin runtime modules are loaded. Startup then stages runtime dependencies only for the

    text
    startupPluginIds
    selected by that plan.

    For packaged installs, dependency staging is allowed before plugin import. After staging, the runtime loader imports startup plugins with install repair disabled; at that point missing dependency materialization is treated as a load failure, not another repair loop.

    When startup dependency staging is deferred behind the HTTP bind, Gateway readiness stays blocked on the

    text
    plugin-runtime-deps
    reason until the selected startup plugin dependencies are materialized and the startup plugin runtime has loaded.

    When repair runs

    Runtime dependency repair should run when one of these is true:

    • the effective plugin plan changed and adds bundled plugins that need runtime dependencies
    • the generated dependency manifest no longer matches the effective plan
    • expected installed package sentinels are missing or incomplete
    • text
      openclaw doctor --fix
      or
      text
      openclaw plugins deps --repair
      was requested

    Runtime dependency repair should not run just because OpenClaw started. A normal startup with an unchanged plan and complete dependency materialization should skip package-manager work.

    Commands that edit config, enable plugins, or repair doctor findings can enter plugin plan mode once, materialize the newly required bundled dependencies, then return to the normal command flow. Local

    text
    openclaw onboard
    and
    text
    openclaw configure
    do this automatically after they successfully write config, so the next Gateway run does not discover missing bundled plugin packages after startup has already begun. Remote onboarding/configure stays read-only for local runtime deps.

    Hot reload rule

    Hot reload paths that can change active plugins must go back through plugin plan mode before loading plugin runtime. The reload should compare the new effective plugin plan with the previous one, stage missing dependencies for newly active bundled plugins, then load or restart the affected runtime.

    If a config reload does not change the effective plugin plan, it should not repair bundled runtime dependencies.

    Package manager execution

    OpenClaw writes a generated install manifest for the selected bundled runtime dependencies and runs the package manager in the runtime dependency install root. It prefers

    text
    pnpm
    when available and falls back to the Node-bundled
    text
    npm
    runner.

    The

    text
    pnpm
    path uses production dependencies, disables lifecycle scripts, ignores the workspace, and keeps the store inside the install root:

    bash
    pnpm install \ --prod \ --ignore-scripts \ --ignore-workspace \ --config.frozen-lockfile=false \ --config.minimum-release-age=0 \ --config.store-dir=<install-root>/.openclaw-pnpm-store \ --config.node-linker=hoisted \ --config.virtual-store-dir=.pnpm

    The

    text
    npm
    fallback uses the safe npm install wrapper with production dependencies, lifecycle scripts disabled, workspace mode disabled, audit disabled, fund output disabled, legacy peer dependency behavior, and package-lock output enabled for the generated install root.

    After install, OpenClaw validates the staged dependency tree before making it visible to the runtime dependency root. Isolated staging is copied into the runtime dependency root and validated again.

    The whole repair/materialization section is guarded by an install-root lock. Current lock owners record PID, process start-time when available, and creation time. Legacy locks without process start-time or creation-time evidence are only reclaimed by filesystem age, so recycled Docker PID 1 locks recover without expiring normal long-running current installs by age alone.

    Install roots

    Packaged installs must not mutate read-only package directories. OpenClaw can read dependency roots from packaged layers, but writes generated runtime dependencies to a writable stage such as:

    • text
      OPENCLAW_PLUGIN_STAGE_DIR
    • text
      $STATE_DIRECTORY
    • text
      ~/.openclaw/plugin-runtime-deps
    • text
      /var/lib/openclaw/plugin-runtime-deps
      in container-style installs

    The writable root is the final materialization target. Older read-only roots are kept as compatibility layers only when needed.

    When a packaged OpenClaw update changes the versioned writable root but the selected bundled-plugin dependency plan is still satisfied by a previous staged root, repair reuses that previous

    text
    node_modules
    tree instead of running the package manager again. The new versioned root still gets its own current package runtime mirror, so plugin code comes from the current OpenClaw package while unchanged dependency trees are shared across updates. Reuse skips previous roots with an active OpenClaw runtime-dependency lock, so a new root does not link to a dependency tree that another Gateway, doctor, or CLI process is currently repairing.

    Doctor and CLI commands

    Use

    text
    plugins deps
    to inspect or repair bundled plugin runtime dependency materialization:

    bash
    openclaw plugins deps openclaw plugins deps --json openclaw plugins deps --repair openclaw plugins deps --prune

    Use doctor when the dependency state is part of broader install health:

    bash
    openclaw doctor openclaw doctor --fix

    text
    plugins deps
    and doctor operate on OpenClaw-owned bundled plugin runtime dependencies selected by the effective plugin plan. They are not third-party plugin install or update commands.

    Troubleshooting

    If a packaged install reports missing bundled runtime dependencies:

    1. Run
      text
      openclaw plugins deps --json
      to inspect the selected plan and missing packages.
    2. Run
      text
      openclaw plugins deps --repair
      or
      text
      openclaw doctor --fix
      to repair the writable dependency stage.
    3. If the install root is read-only, set
      text
      OPENCLAW_PLUGIN_STAGE_DIR
      to a writable path and rerun repair.
    4. Restart Gateway after repair if the missing dependency blocked startup plugin loading.

    In source checkouts, the workspace install usually provides bundled plugin dependencies. Run

    text
    pnpm install
    for source dependency repair instead of using packaged runtime dependency repair as the first step.

    © 2024 TaskFlow Mirror

    Powered by TaskFlow Sync Engine