Skip to content

MAINTAINERS.md

Latest commit

 

History

History
71 lines (54 loc) · 2.63 KB

MAINTAINERS.md

File metadata and controls

71 lines (54 loc) · 2.63 KB

Maintainers Guide

This repo is maintained by agents. This guide documents the infrastructure setup.

Repository Secrets Required

Secret Purpose Status
ANTHROPIC_API_KEY Agent review pipeline (agent-review.yml, claude.yml) Required
ELIZAOS_CLOUD_API_KEY Cloud live E2E tests Optional
ELIZAOS_CLOUD_BASE_URL Cloud live E2E tests Optional

Recommended Branch Protection (main)

Enable these rules on the main branch:

  • Require pull request before merging — no direct pushes
  • Required status checks:
    • lint (from ci.yml)
    • typecheck (from ci.yml)
    • test (from ci.yml)
    • build (from ci.yml)
    • review-pr (from agent-review.yml)
  • Require branches to be up to date before merging
  • Do not allow bypassing the above settings (even for admins)

CI Workflows

Workflow Trigger Purpose
agent-review.yml PRs + Issues Semantic agent review, scope check, security, dark forest
ci.yml PRs + Push to main/develop Lint, typecheck, unit tests, build
test.yml PRs + Push to main/develop Full test suite (unit, e2e, electron, cloud, validation)
auto-label.yml PRs + Issues Auto-label based on file paths and content
claude.yml Comment mentions @claude mentions in PR/issue comments

Agent Review Behavior

The agent reviewer (agent-review.yml) uses these reference docs:

  • SCOPE.md — what's in scope vs out of scope
  • AGENTS.md — coding standards and review priorities
  • CONTRIBUTING.md — contributor agreement and roles

Review Decisions

  • APPROVE — in scope, code quality good, tests adequate, security clear
  • REQUEST CHANGES — fixable issues found
  • CLOSE — out of scope (aesthetic changes, scope creep)

Issue Triage

  • Valid bugs → labeled and kept open
  • QA reports → labeled and prioritized
  • Aesthetic/feature requests → closed with explanation
  • Vague reports → asked for more info

Human Roles

Role Can Do Cannot Do
QA Tester File bug reports, QA reports Submit code, review PRs
Admin (Shaw/Shadow) Merge PRs, configure repo Override agent review decisions*

*The goal is to reach a state where even admins don't need to intervene. Agent decisions should be trusted.

Future Additions

  • Trust scoring for contributors (plugin-trust integration)
  • Auto-merge for agent-approved + all-checks-passing PRs
  • Benchmark gates for performance-sensitive changes
  • Coding agent spawning for issue resolution
  • Reviewer agent separate from coding agent