mirror of https://github.com/stephenleo/bmad-autonomous-development.git synced 2026-04-25 20:34:55 +02:00

Files

stephenleo 303af472f9 feat: initial release of BAD — BMad Autonomous Development v1.0.0

Standalone BMad module that orchestrates fully autonomous parallel
multi-agent pipelines through the full story lifecycle (create → dev →
review → PR), driven by the sprint backlog and dependency graph.

- Every step runs in a dedicated subagent with a fresh context window
- Harness-agnostic: detects Claude Code, Cursor, Copilot, etc. at setup
- Configurable models, timers, CI, and merge behaviour per harness
- Self-registering via assets/module-setup.md

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

2026-04-05 12:49:27 +08:00

2.9 KiB

Raw Blame History

Phase 0: Dependency Graph — Detailed Reference

Read this file during Phase 0 steps 4 and 6.

Step 4: Update GitHub PR/Issue Status and Reconcile sprint-status.yaml

GitHub PR merge status is the authoritative source of truth for whether a story is done. This step always runs, even on resume.

PR Status Lookup

Search by branch name:

gh pr list --search "story-{number}" --state all --json number,title,state,mergedAt

GitHub Issue Number Lookup

Resolve in this order:

Check the epic file and sprint-status.yaml for an explicit issue reference.
If not found, search by the BMad issue title prefix "Story {number}:":
```
gh issue list --search "Story 3.1:" --json number,title,state
```
Pick the best match by comparing titles.
If still not found, leave the Issue column blank.

Reconcile sprint-status.yaml from PR Status

After updating PR statuses, sync _bmad-output/implementation-artifacts/sprint-status.yaml at the repo root to match. For every story whose PR status is now merged, set its sprint-status entry to done (regardless of what the file currently says).

This repair step handles cases where sprint-status.yaml was reset or reverted by a git operation — GitHub is always right.

Step 6: Dependency Graph Format

Write from scratch on first run. On subsequent runs, update only columns that change (Sprint Status, Issue, PR, PR Status, Ready to Work), add new rows for new stories, and preserve all existing dependency chain data.

Schema

# Story Dependency Graph
_Last updated: {ISO timestamp}_

## Stories

| Story | Epic | Title | Sprint Status | Issue | PR | PR Status | Dependencies | Ready to Work |
|-------|------|-------|--------------|-------|----|-----------|--------------|---------------|
| 1.1   | 1    | ...   | done         | #10   | #42 | merged  | none         | ✅ Yes (done) |
| 1.2   | 1    | ...   | backlog      | #11   | #43 | open    | 1.1          | ❌ No (1.1 not merged) |
| 1.3   | 1    | ...   | backlog      | #12   | —   | —       | none         | ✅ Yes        |
| 2.1   | 2    | ...   | backlog      | #13   | —   | —       | none         | ❌ No (epic 1 not complete) |

## Dependency Chains

- **1.2** depends on: 1.1
- **1.4** depends on: 1.2, 1.3
...

## Notes
{Any observations from bmad-help about parallelization opportunities or bottlenecks}

Ready to Work Rules

Ready to Work = ✅ Yes only when all of the following are true:

The story itself is not done.
Every story it depends on has a merged PR (or is done with a merged PR).
Every story in all lower-numbered epics has a merged PR (or is done with a merged PR) — epic N may not start until epic N-1 is fully merged into main.

Any condition failing → ❌ No with a parenthetical explaining the blocker (e.g., ❌ No (1.1 not merged), ❌ No (epic 1 not complete)).

2.9 KiB Raw Blame History