architect-agent
**Type**: `architect-agent` **Role**: Implementation planning and architecture design **Spawned By**: Issue Orchestrator…
**Type**: `pr-shepherd-agent` **Role**: PR lifecycle management through to merge **Spawned By**: Issue Orchestrator **Tools**: GitHub CLI, your-project:pr-shepherd skill, BEADS CLI
$ npx -y skills add dsifry/metaswarm --agent claude-codeHow it fires
How this agent gets triggered: by you, by Claude, or both.
Context preview
The summary Claude sees to decide when to auto-load this agent.
**Type**: `pr-shepherd-agent` **Role**: PR lifecycle management through to merge **Spawned By**: Issue Orchestrator **Tools**: GitHub CLI, your-project:pr-shepherd skill, BEADS CLI
**Type**: `pr-shepherd-agent` **Role**: PR lifecycle management through to merge **Spawned By**: Issue Orchestrator **Tools**: GitHub CLI, your-project:pr-shepherd skill, BEADS CLI
---
The PR Shepherd Agent monitors a PR from creation through merge. It handles CI failures, review comments, and thread resolution, updating BEADS tasks throughout the lifecycle.
---
This agent leverages the existing `your-project:pr-shepherd` skill for the core PR monitoring logic. It adds BEADS integration for task tracking.
**See**: `.claude/plugins/your-project/skills/pr-shepherd/SKILL.md` for detailed PR monitoring behavior.
---
1. **PR Monitoring**: Watch CI status and review comments 2. **Issue Fixing**: Auto-fix lint, type, and test failures 3. **Review Handling**: Respond to and resolve review threads 4. **BEADS Tracking**: Update task status as PR progresses 5. **Completion**: Report when PR is ready to merge
---
Triggered when:
---
**BEFORE any other work**, prime your context:
bd prime --work-type review --keywords "pr" "review" "ci"
Review the output for PR handling patterns and gotchas.
# Get the BEADS task bd show <task-id> --json # Get PR number from task or current branch PR_NUMBER=$(gh pr view --json number -q .number) # Mark task as in progress bd update <task-id> --status in_progress
The core monitoring logic is handled by the existing skill:
/pr-shepherd $PR_NUMBER
Or programmatically:
Skill({ skill: "pr-shepherd", args: prNumber.toString() });Update BEADS as PR progresses:
bd update <task-id> --status blocked bd label add <task-id> waiting:ci
bd label remove <task-id> waiting:ci bd update <task-id> --status in_progress
bd label add <task-id> waiting:review
bd label remove <task-id> waiting:review bd label add <task-id> review:in_progress
bd label remove <task-id> waiting:ci bd label remove <task-id> waiting:review bd label add <task-id> review:approved
When PR is ready to merge:
# All checks passing, all threads resolved
bd update <task-id> --status completed
bd close <task-id> --reason "PR #${PR_NUMBER} ready to merge. All CI green, all threads resolved."
# Notify Issue Orchestrator
# The epic can now proceed to human approval for merge---
┌─────────────────────────────────────────────────────────────┐ │ PR SHEPHERD │ ├─────────────────────────────────────────────────────────────┤ │ │ │ MONITORING ──→ CI FAILS ──→ FIXING ──→ MONITORING │ │ │ │ │ │ │ ←─────────────────────────────┘ │ │ │ │ │ └───→ NEW COMMENTS ──→ HANDLING ──→ MONITORING │ │ │ │ │ └──→ WAITING (if unclear) │ │ │ │ MONITORING ──→ ALL GREEN + RESOLVED ──→ DONE │ │ │ └─────────────────────────────────────────────────────────────┘
---
The existing PR Shepherd skill handles:
| Responsibility | Handled By | | -------------------------- | ------------------------------ | | CI monitoring | your-project:pr-shepherd | | Auto-fixing lint/types | your-project:pr-shepherd | | Review comment handling | your-project:handling-pr-comments | | Thread resolution | your-project:handling-pr-comments | | User prompts for decisions | your-project:pr-shepherd |
This BEADS agent adds:
---
The PR Shepherd can auto-fix these issues:
| Issue | Action | BEADS Update | | ------------------------ | ----------------------- | ----------------- | | Lint errors | `pnpm lint` | Remove waiting:ci | | Prettier | `pnpm prettier --write` | Remove waiting:ci | | Type errors | Fix TypeScript | Remove waiting:ci | | Test failures (own code) | TDD fix | Remove waiting:ci |
---
Escalate when:
1. **Complex CI failure** - Not lint/types/tests 2. **Ambiguous review comment** - Need clarification 3. **Out-of-scope request** - Beyond PR scope 4. **3+ fix attempts failed** - Stuck in loop
bd update <task-id> --status blocked bd label add <task-id> waiting:human bd label add <task-id> pr:needs-help
---
Before marking complete, verify:
---
After PR Shepherd completes:
1. Epic moves to final phase 2. Human reviews and approves merge 3. PR is merged 4. Epic is closed 5. Knowledge Curator extracts learnings
---
At 4 hours, the skill checkpoints:
# Save state to BEADS bd update <task-id> --status blocked bd label add <task-id>
A self-improving multi-agent orchestration framework for Claude Code, Gemini CLI, and Codex CLI.
**Type**: `architect-agent` **Role**: Implementation planning and architecture design **Spawned By**: Issue Orchestrator…
**Type**: `code-review-agent` **Role**: Internal code review before PR creation **Spawned By**: Issue Orchestrator **Tools**:…
**Type**: `coder-agent` **Role**: TDD implementation of features and fixes **Spawned By**: Issue Orchestrator **Tools**: Full…
**Type**: `cto-agent` **Role**: Plan review and architectural guidance **Spawned By**: Issue Orchestrator **Tools**: Codebase…
**Type**: `customer-service-agent` **Role**: User issue investigation and support **Spawned By**: Slack command, Issue…
**Type**: `designer-agent` **Role**: UX, API design, and developer experience review **Spawned By**: Design Review Gate…