Run BMAD Autopilot autonomous development
/plugin marketplace add hanibalsk/claude-marketplace/plugin install bmad-system@hanibalsk-marketplaceRun autonomous development following the BMAD Method with git worktree isolation.
/bmad [epic-pattern] [options]
epic-pattern - Optional. Epic IDs to process (e.g., "7A", "7A 8A", "10.*")
--worktree - Create isolated git worktree for development# Process all epics
/bmad
# Process specific epic with worktree isolation
/bmad 7A --worktree
# Process multiple epics in parallel (each gets own worktree)
/bmad "7A 8A 10B" --worktree
# Process epics matching pattern
/bmad "10.*"
--worktree)feature/epic-{id}FIND_EPIC → CREATE_BRANCH → DEVELOP_STORIES → CODE_REVIEW
↓
DONE ← MERGE_PR ← WAIT_COPILOT ← CREATE_PR ← (approved)
↓
FIX_ISSUES
BMAD uses specialized agent threads:
| Agent | Role |
|---|---|
| Coordinator | Finds epics, manages workflow |
| Developer | Implements stories |
| Reviewer | Code review |
| PR Manager | Creates/manages PRs |
| Fixer | Fixes CI/review issues |
| Monitor | Watches PR status |
# Check current status
ct thread status bmad-main
# Follow logs
ct thread logs bmad-main -f
# View events
ct event list --type "EPIC_*"
# List active worktrees
ct worktree list
Set in .claude-threads/config.yaml:
bmad:
epic_pattern: ""
base_branch: main
auto_merge: true
max_concurrent_prs: 2
check_interval: 300
worktrees:
enabled: true
max_age_days: 7
auto_cleanup: true
default_base_branch: main
auto_push: true
pr_shepherd:
max_fix_attempts: 5
ci_poll_interval: 30
auto_merge: false
If autopilot gets stuck:
# Check why
ct thread status bmad-main --verbose
# Resume
ct thread resume bmad-main
# Or restart
ct thread stop bmad-main
ct thread start bmad-main
BMAD expects stories in:
_bmad-output/stories/epic-{id}/docs/stories/{id}/Watch the development progress:
# Real-time events
ct event subscribe "STORY_*,PR_*"
# Recent events
ct event list --limit 50
Enable webhooks for real-time PR updates:
ct webhook start --port 31338
# Configure in GitHub: http://server:31338/webhook
# Start fresh sprint
/bmad "sprint-12.*"
# Monitor progress
ct thread list running
# Check for issues
ct event list --type "*_BLOCKED,*_FAILED"
# Develop one epic interactively
ct thread create epic-7a \
--mode interactive \
--template bmad-developer.md \
--context '{"epic_id": "7A"}'
ct thread resume epic-7a
# Process multiple epics concurrently with isolated worktrees
ct thread create epic-7a --mode automatic --template bmad-developer.md --worktree --context '{"epic_id": "7A"}'
ct thread create epic-8a --mode automatic --template bmad-developer.md --worktree --context '{"epic_id": "8A"}'
ct thread create epic-9a --mode automatic --template bmad-developer.md --worktree --context '{"epic_id": "9A"}'
# Each epic runs in its own isolated worktree
ct orchestrator start
# Monitor all worktrees
ct worktree list
# Watch PRs with automatic fix handling
ct pr watch 123
# Shepherd creates isolated worktree for fixes
# Automatically fixes CI failures and review comments
ct pr status 123
# Run as background daemon
ct pr daemon