# Molecule AI Dev Team — PM + Research + Dev
name: Molecule AI Dev Team
description: AI agent company for building Molecule AI

defaults:
  runtime: claude-code
  tier: 2
  # required_env removed — PR #1031 eliminated the CLAUDE_CODE_OAUTH_TOKEN
  # requirement; workspaces authenticate via the platform token flow.
  # ANTHROPIC_API_KEY is set via workspace .env, not as a required_env constraint.

  # Default plugin set applied to every workspace. Per-workspace `plugins:`
  # UNIONs with this set (#71). Use just the additions; prefix `!` (or `-`)
  # to opt a default OUT for one workspace if needed.
  #
  # Coding / guardrail essentials:
  # - ecc:                            "Everything Claude Code" guardrails + coding skills
  # - molecule-dev:                   Molecule AI codebase conventions, past bugs, review-loop
  # - superpowers:                    systematic-debugging, TDD, planning, verification-before-completion
  #
  # Safety hooks (PreToolUse/PostToolUse/UserPromptSubmit) — universal:
  # - molecule-careful-bash:          refuse destructive shell (rm -rf, push --force main, DROP TABLE)
  # - molecule-prompt-watchdog:       inject warnings on destructive user prompts
  # - molecule-audit-trail:           append every Edit/Write to .claude/audit.jsonl
  #
  # Operational memory — keeps agents consistent across sessions/cron ticks:
  # - molecule-session-context:       auto-load cron learnings + PR/issue counts on SessionStart
  # - molecule-skill-cron-learnings:  per-tick learning JSONL format (pairs with session-context)
  #
  # Docs hygiene:
  # - molecule-skill-update-docs:     keep architecture / README / edit-history aligned with code
  plugins:
    - ecc
    - molecule-dev
    - superpowers
    - molecule-careful-bash
    - molecule-prompt-watchdog
    - molecule-audit-trail
    - molecule-session-context
    - molecule-skill-cron-learnings
    - molecule-skill-update-docs

  # Audit-summary routing — generic per-template mapping (issue #51).
  # Auditors (Security Auditor, UIUX Designer, QA Engineer) send A2A messages
  # with metadata.audit_summary.category set. The receiver (PM) reads this
  # table from its own /configs/config.yaml and delegates to each listed role.
  # Each org template owns its own mapping — role names are NOT hardcoded in
  # prompts, so adding/renaming roles is a config-only change.
  category_routing:
    # Defensive findings — code review, SAST, missing patch class.
    security: [Backend Engineer, Backend Engineer 2, Backend Engineer 3, DevOps Engineer]
    # Adversarial findings — live exploit, container escape, supply-chain
    # CVE, cross-agent prompt injection. Routed through Security Auditor
    # first so the patch class is named before paging the implementing
    # engineer; most red-team findings need both an immediate mitigation
    # (DevOps) and a structural fix (Security Auditor + Backend).
    offensive: [Security Auditor, Security Auditor 2, Backend Engineer, DevOps Engineer]
    ui: [Frontend Engineer, Frontend Engineer 2, Frontend Engineer 3]
    ux: [Frontend Engineer, Frontend Engineer 2]
    infra: [DevOps Engineer, Platform Engineer, SRE Engineer]
    # Cloud-services findings (Vercel/Fly/GHCR/Upptime) — failed deploy,
    # broken health check, expired cert, increased error rate, runaway
    # cost. Routed to DevOps first (operate the platform) and Backend
    # second (most cloud regressions trace back to a backend deploy).
    cloud: [DevOps Engineer, Platform Engineer, SRE Engineer, Backend Engineer]
    qa: [QA Engineer, QA Engineer 2, QA Engineer 3]
    performance: [Backend Engineer]
    docs: [Documentation Specialist]
    mixed: [Dev Lead]
    # Evolution-cron categories (#93): these four are fired by hourly
    # self-review schedules (Research Lead, Technical Researcher, Dev Lead,
    # DevOps Engineer). Routing them to the same role that generated them
    # is a safe default — it converts the summary into a delegation back
    # to the author so they act on their own findings. Override per-org
    # if you want a different fan-out.
    research: [Research Lead]
    plugins: [Technical Researcher]
    template: [Dev Lead]
    channels: [DevOps Engineer]
    # Marketing team categories (2026-04-16). Peer sub-tree under CEO —
    # reports via Marketing Lead for coordination + cross-functional
    # delegations into the dev team (DevRel → Backend Engineer for code
    # samples, PMM → Competitive Intelligence for eco-watch diffs).
    content: [Content Marketer]
    positioning: [Product Marketing Manager]
    community: [Community Manager]
    growth: [SEO Growth Analyst]
    social: [Social Media Brand]
    devrel: [DevRel Engineer]

  # workspace_dir: not set by default — each agent gets an isolated Docker volume
  # Set per-workspace to bind-mount a host directory as /workspace

  # Idle-loop reflection pattern (#205). When idle_prompt is non-empty, the
  # workspace self-sends this prompt every idle_interval_seconds while its
  # heartbeat.active_tasks == 0. Pattern from Hermes/Letta. Cost collapses to
  # event-driven (no LLM call unless there's actually nothing to do). Off by
  # default to avoid surprising token burn — set per-workspace to enable.
  # Keep idle prompts local (no A2A sends): same rule as initial_prompt.
  idle_prompt: ""
  idle_interval_seconds: 600   # 10 min — ignored when idle_prompt is empty

  # initial_prompt runs once on first boot (not on restart).
  # Molecule-AI/molecule-core is a container env var from .env secrets.
  # IMPORTANT: Do NOT send A2A messages in initial_prompt — other agents may not
  # be ready yet. Keep it local: clone, read, memorize. Wait for tasks.
  initial_prompt: |
    You just started. Set up your environment silently — do NOT contact other agents yet.
    SCOPE: this team owns the entire Molecule-AI GitHub org (40+ repos: molecule-core,
    molecule-app, docs, landingpage, all plugin/template/sdk repos) AND the live cloud
    services that run them (Vercel for app + landingpage, Fly for control plane, GHCR
    for images, Upptime for status). Do NOT scope yourselves to molecule-core only —
    issues, PRs, and incidents from any Molecule-AI/* repo are in scope. Pick up work
    from across the org based on your role.

    1. Clone the baseline repo (molecule-core conventions are the org standard).
       Authenticated when GITEA_TOKEN is available, anonymous otherwise. When a token
       is present, use it in-URL ONLY for the clone, then scrub the remote URL so the
       token is never persisted to /workspace/repos/molecule-core/.git/config:
       mkdir -p /workspace/repos
       if [ -n "$GITEA_TOKEN" ]; then
         git clone "https://x-access-token:${GITEA_TOKEN}@git.moleculesai.app/molecule-ai/molecule-core.git" /workspace/repos/molecule-core 2>/dev/null \
           && (cd /workspace/repos/molecule-core && git remote set-url origin "https://git.moleculesai.app/molecule-ai/molecule-core.git") \
           || (cd /workspace/repos/molecule-core && git pull)
       else
         git clone "https://git.moleculesai.app/molecule-ai/molecule-core.git" /workspace/repos/molecule-core 2>/dev/null || (cd /workspace/repos/molecule-core && git pull)
       fi
       # Backwards-compat symlink — older role prompts still reference /workspace/repo
       ln -sfn /workspace/repos/molecule-core /workspace/repo

    2. Enumerate the org so you know what's out there. Don't clone everything (wasteful);
       memorise the inventory and clone-on-demand when a task touches a specific repo:
         gh repo list Molecule-AI --limit 60 --json name,description,updatedAt \
           > /workspace/org-repos.json
       Use commit_memory with key `org-repos-inventory` to save the repo list +
       brief description for each. Re-enumerate on every restart so you stay current
       with new repos.

    3. Set up git hooks for the baseline: cd /workspace/repos/molecule-core && git config core.hooksPath .githooks
    4. Read /workspace/repos/molecule-core/CLAUDE.md to understand the project conventions.
    5. Read your system prompt at /configs/system-prompt.md to understand your role + which
       org repos and cloud services are YOUR ownership area.
    6. Save key conventions to memory so you recall them on every future task:
       Use commit_memory to save: "CONVENTIONS: (1) Every canvas .tsx using hooks needs 'use client' as first line — run the grep check before committing. (2) Dark zinc theme only — never white/light. (3) Zustand selectors must not create new objects. (4) Always run npm test + npm run build before reporting done. (5) Use delegate_task to ask peers questions directly — don't guess API shapes. (6) Pre-commit hook at .githooks/pre-commit enforces these — commits will be rejected if violated. (7) When working in a non-core repo, clone it on demand under /workspace/repos/<name> and follow that repo's CLAUDE.md if present."
    7. You are now ready. Wait for tasks from your parent — do not initiate contact.

workspaces:
  - !include teams/pm.yaml
  - !include teams/marketing.yaml
  # Dev tree composed via gitops `!external` resolver (internal#77 / molecule-core#105).
  # Platform fetches molecule-ai/molecule-dev-department at import time into a
  # content-addressable cache under <orgBaseDir>/.external-cache/. No operator-
  # side filesystem symlink, no sibling-clone deploy contract — the import is
  # self-contained. Pinned to v1.0.0 (first stable atomized release;
  # see molecule-ai/molecule-dev-department tags for newer versions).
  - !external
    repo: molecule-ai/molecule-dev-department
    ref: v1.0.0
    path: dev-lead/workspace.yaml

template_schema_version: 1