Config Schema

Kino-configure ang Triggerfish sa pamamagitan ng triggerfish.yaml, na matatagpuan sa ~/.triggerfish/triggerfish.yaml pagkatapos patakbuhin ang triggerfish dive. Dino-document ng page na ito ang bawat configuration section.

Mga Secret Reference Anumang string value sa file na ito ay maaaring gumamit ng secret: prefix para mag-reference ng credential na naka-store sa OS keychain. Halimbawa, ang apiKey: "secret:provider:anthropic:apiKey" ay nire-resolve ang value mula sa keychain sa startup. Tingnan ang Secrets Management para sa mga detalye. :::

Buong Annotated Example

# =============================================================================
# triggerfish.yaml -- Complete configuration reference
# =============================================================================

# ---------------------------------------------------------------------------
# Models: LLM provider configuration at failover
# ---------------------------------------------------------------------------
models:
  # Ang primary model na ginagamit para sa agent completions
  primary:
    provider: anthropic
    model: claude-sonnet-4-5

  # Optional: hiwalay na vision model para sa image description
  # Kapag hindi sumusuporta ng vision ang primary model, awtomatikong
  # dine-describe ng model na ito ang mga images bago ma-reach ang primary.
  # vision: glm-4.5v

  # Streaming responses (default: true)
  # streaming: true

  # Provider-specific configuration
  # Ang mga API keys ay nire-reference sa pamamagitan ng secret: syntax at nire-resolve mula sa OS keychain.
  # Patakbuhin ang `triggerfish dive` o `triggerfish config migrate-secrets` para i-set up.
  providers:
    anthropic:
      model: claude-sonnet-4-5
      # apiKey: "secret:provider:anthropic:apiKey"

    openai:
      model: gpt-4o

    google:
      model: gemini-pro

    ollama:
      model: llama3
      endpoint: "http://localhost:11434"

    lmstudio:
      model: lmstudio-community/Meta-Llama-3.1-8B-Instruct-GGUF
      endpoint: "http://localhost:1234"

    openrouter:
      model: anthropic/claude-sonnet-4-5

    zenmux:
      model: openai/gpt-5

    zai:
      model: glm-4.7

  # Ordered failover chain -- sinusubukan sa pagkakasunod-sunod kapag nag-fail ang primary
  failover:
    - claude-haiku-4-5 # Unang fallback
    - gpt-4o # Pangalawang fallback
    - ollama/llama3 # Local fallback (hindi kailangan ng internet)

  # Failover behavior
  failover_config:
    max_retries: 3 # Mga retry bawat provider bago lumipat sa susunod
    retry_delay_ms: 1000 # Delay sa pagitan ng retries
    conditions: # Ano ang nagti-trigger ng failover
      - rate_limited # Nagbalik ang provider ng 429
      - server_error # Nagbalik ang provider ng 5xx
      - timeout # Lumampas sa timeout ang request

# ---------------------------------------------------------------------------
# Logging: Structured log output
# ---------------------------------------------------------------------------
logging:
  level: normal # quiet | normal | verbose | debug

# ---------------------------------------------------------------------------
# Channels: Mga messaging platform connection
# ---------------------------------------------------------------------------
# Ang mga secrets (bot tokens, API keys, passwords) ay naka-store sa OS keychain.
# Patakbuhin ang `triggerfish config add-channel <name>` para ma-enter nang ligtas.
# Ang non-secret configuration lang ang lumalabas dito.
channels:
  telegram:
    ownerId: 123456789 # Iyong Telegram numeric user ID
    classification: INTERNAL # Default: INTERNAL

  signal:
    endpoint: "tcp://127.0.0.1:7583" # signal-cli daemon endpoint
    account: "+14155552671" # Iyong Signal phone number (E.164)
    classification: PUBLIC # Default: PUBLIC
    defaultGroupMode: mentioned-only # always | mentioned-only | owner-only
    groups:
      "group-id-here":
        mode: always
        classification: INTERNAL

  slack:
    classification: PUBLIC # Default: PUBLIC

  discord:
    ownerId: "your-discord-user-id" # Iyong Discord user ID
    classification: PUBLIC # Default: PUBLIC

  whatsapp:
    phoneNumberId: "your-phone-number-id" # Mula sa Meta Business Dashboard
    classification: PUBLIC # Default: PUBLIC

  webchat:
    port: 8765 # WebSocket port para sa web client
    classification: PUBLIC # Default: PUBLIC (visitors)

  email:
    smtpApiUrl: "https://api.sendgrid.com/v3/mail/send"
    imapHost: "imap.gmail.com"
    imapPort: 993
    imapUser: "you@gmail.com"
    fromAddress: "bot@example.com"
    ownerEmail: "you@gmail.com"
    classification: CONFIDENTIAL # Default: CONFIDENTIAL

# ---------------------------------------------------------------------------
# Classification: Data sensitivity model
# ---------------------------------------------------------------------------
classification:
  mode: personal # "personal" o "enterprise" (coming soon)
# Levels: RESTRICTED > CONFIDENTIAL > INTERNAL > PUBLIC

# ---------------------------------------------------------------------------
# Policy: Custom enforcement rules (enterprise escape hatch)
# ---------------------------------------------------------------------------
policy:
  rules:
    - id: block-external-pii
      hook: PRE_OUTPUT
      priority: 100
      conditions:
        - type: recipient_is
          value: EXTERNAL
        - type: content_matches
          pattern: "\\b\\d{3}-\\d{2}-\\d{4}\\b" # SSN pattern
      action: REDACT
      message: "PII redacted for external recipient"

    - id: rate-limit-browser
      hook: PRE_TOOL_CALL
      priority: 50
      conditions:
        - type: tool_name
          value: browser
        - type: rate_exceeds
          value: 10/minute
      action: BLOCK
      message: "Browser tool rate limit exceeded"

# ---------------------------------------------------------------------------
# MCP Servers: External tool servers
# ---------------------------------------------------------------------------
mcp_servers:
  filesystem:
    command: "deno"
    args: ["run", "--allow-read", "--allow-write", "mcp-filesystem-server.ts"]
    classification: INTERNAL

  github:
    command: "npx"
    args: ["-y", "@modelcontextprotocol/server-github"]
    classification: CONFIDENTIAL

# ---------------------------------------------------------------------------
# Scheduler: Cron jobs at triggers
# ---------------------------------------------------------------------------
scheduler:
  cron:
    jobs:
      - id: morning-briefing
        schedule: "0 7 * * *" # 7 AM araw-araw
        task: "Prepare morning briefing with calendar, unread emails, and weather"
        channel: telegram
        classification: INTERNAL

      - id: pipeline-check
        schedule: "0 */4 * * *" # Tuwing 4 oras
        task: "Check Salesforce pipeline for changes and notify if significant"
        channel: slack
        classification: CONFIDENTIAL

      - id: pr-review-check
        schedule: "*/15 * * * *" # Tuwing 15 minuto
        task: "Check open PR tracking files and query GitHub for new reviews"
        classification: INTERNAL

  trigger:
    interval: 30m # I-check tuwing 30 minuto
    classification: INTERNAL # Max taint ceiling para sa triggers
    quiet_hours: "22:00-07:00" # I-suppress sa mga oras na ito

# ---------------------------------------------------------------------------
# Notifications: Mga delivery preference
# ---------------------------------------------------------------------------
notifications:
  preferred_channel: telegram # Default delivery channel
  quiet_hours: "22:00-07:00" # I-suppress ang normal/low priority
  batch_interval: 15m # I-batch ang low-priority notifications

# ---------------------------------------------------------------------------
# Agents: Multi-agent routing (optional)
# ---------------------------------------------------------------------------
agents:
  default: personal # Fallback agent
  list:
    - id: personal
      name: "Personal Assistant"
      channels: [whatsapp, telegram]
      tools:
        profile: "full"
      model: claude-opus-4-5
      classification_ceiling: INTERNAL

    - id: work
      name: "Work Assistant"
      channels: [slack, email]
      tools:
        profile: "coding"
        allow: [browser, github]
      model: claude-sonnet-4-5
      classification_ceiling: CONFIDENTIAL

# ---------------------------------------------------------------------------
# Voice: Speech configuration (optional)
# ---------------------------------------------------------------------------
voice:
  stt:
    provider: whisper # whisper | deepgram | openai
    model: base # Whisper model size
  tts:
    provider: elevenlabs # elevenlabs | openai | system
    voice_id: "your-voice-id"
  wake_word: "triggerfish"
  push_to_talk:
    shortcut: "Ctrl+Space"

# ---------------------------------------------------------------------------
# Webhooks: Inbound event endpoints (optional)
# ---------------------------------------------------------------------------
webhooks:
  endpoints:
    - id: github
      path: /webhook/github
      # Ang webhook secret ay naka-store sa OS keychain
      classification: INTERNAL
      actions:
        - event: "pull_request.opened"
          task: "Review PR and post summary"
        - event: "pull_request_review"
          task: "A PR review was submitted. Read tracking file, address feedback, commit, push."
        - event: "pull_request_review_comment"
          task: "An inline review comment was posted. Read tracking file, address comment."
        - event: "issue_comment"
          task: "A comment was posted on a PR. If tracked, address feedback."
        - event: "pull_request.closed"
          task: "PR closed or merged. Clean up branches and archive tracking file."
        - event: "issues.opened"
          task: "Triage new issue"

# ---------------------------------------------------------------------------
# GitHub: Mga GitHub integration setting (optional)
# ---------------------------------------------------------------------------
github:
  auto_merge: false # Default: false. I-set sa true para auto-merge ng approved PRs.

# ---------------------------------------------------------------------------
# Groups: Group chat behavior (optional)
# ---------------------------------------------------------------------------
groups:
  default_behavior: "mentioned-only"
  overrides:
    - channel: slack
      channel_name: "#ai-assistant"
      behavior: "always"

# ---------------------------------------------------------------------------
# Web: Search at fetch configuration
# ---------------------------------------------------------------------------
web:
  search:
    provider: brave # Search backend (brave ang default)
# Ang API key ay naka-store sa OS keychain

# ---------------------------------------------------------------------------
# Remote: Remote access (optional)
# ---------------------------------------------------------------------------
remote:
  tailscale:
    serve: true
    funnel:
      enabled: true
      paths: ["/webhook/*"]
  auth:
# Ang auth token ay naka-store sa OS keychain

Section Reference

models

Key	Type	Paglalarawan
primary	object	Primary model reference na may provider at model fields
primary.provider	string	Provider name (anthropic, openai, google, ollama, lmstudio, openrouter, zenmux, zai)
primary.model	string	Model identifier na ginagamit para sa agent completions
vision	string	Optional vision model para sa automatic image description (tingnan ang Image and Vision)
streaming	boolean	I-enable ang streaming responses (default: true)
providers	object	Provider-specific configuration (tingnan sa ibaba)
failover	string[]	Ordered list ng fallback models
failover_config.max_retries	number	Mga retry bawat provider bago mag-failover
failover_config.retry_delay_ms	number	Delay sa pagitan ng retries sa milliseconds
failover_config.conditions	string[]	Mga conditions na nagti-trigger ng failover

channels

Bawat channel key ay ang channel type. Lahat ng channel types ay sumusuporta ng classification field para i-override ang default classification level.

Lahat ng secrets (tokens, API keys, passwords) ay naka-store sa OS keychain, hindi sa file na ito. Patakbuhin ang triggerfish config add-channel <name> para ma-enter ang credentials nang ligtas. :::

classification

Key	Type	Paglalarawan
mode	"personal" o "enterprise"	Deployment mode (coming soon -- kasalukuyang parehong gumagamit ng parehong classification levels)

policy

Custom rules na ine-evaluate sa hook execution. Bawat rule ay nagsi-specify ng hook type, priority, conditions, at action. Ang mas mataas na priority numbers ay ine-evaluate muna.

mcp_servers

External MCP tool servers. Bawat server ay nagsi-specify ng command para i-launch ito, optional environment variables, classification level, at per-tool permissions.

scheduler

Cron job definitions at trigger timing. Tingnan ang Cron and Triggers para sa mga detalye.

notifications

Notification delivery preferences. Tingnan ang Notifications para sa mga detalye.

web

Key	Type	Paglalarawan
web.search.provider	string	Search backend para sa web_search tool (kasalukuyan: brave)

Tingnan ang Web Search and Fetch para sa mga detalye.

logging

Key	Type	Default	Paglalarawan
level	string	"normal"	Log verbosity: quiet (errors lang), normal (info), verbose (debug), debug (trace)

Tingnan ang Structured Logging para sa mga detalye tungkol sa log output at file rotation.

github

Key	Type	Default	Paglalarawan
auto_merge	boolean	false	Kapag true, auto-merge ng agent ang PRs pagkatapos makatanggap ng approving review. Kapag false (default), ino-notify ng agent ang owner at naghihintay ng explicit merge instruction.

Tingnan ang GitHub Integration guide para sa buong setup instructions.