status

You are printing a status dashboard for the current brigade state. Read-only — never modify files. Goal: give the user a quick overview without opening 10 files manually.

Step 1 — Locate state

Check if this is a brigade-enabled project:

• If .planning/tasks/ROLES.md does not exist → print: "Not a brigade project. Run /brigade:init to bootstrap." and stop.

Read:

• .planning/tasks/ROLES.md — roles and agent mapping
• .planning/ARCHITECTURE.md (if exists) — module count
• .planning/learnings.md (if exists) — recent patterns
• .planning/tasks/sprint-*/ — all sprint directories

Step 2 — Identify current sprint

Pick the highest-numbered sprint. Read its _README.md for goal, waves, file map.

Step 3 — Parse task specs

For each task file in the current sprint (S{N}.{M}-*.md):

1. Read frontmatter: id, title, role, status, wave, depends
2. Determine effective status:
   • done — status: done
   • in-progress — status: in-progress OR has feature branch feat/s{N}.{M}-* in git
   • blocked — any depends task is not yet done
   • todo — otherwise

Step 4 — Check review reports

For each wave with reports in reviews/:

• Read wave-{W}-summary.md if exists → extract verdict + counts
• Otherwise scan individual review files

Step 5 — Print the status

Format (use ANSI colors for terminal, fall back to plain text if not):

brigade status — {project-path}

Sprint {N} — {done}/{total} tasks done ({pct}%)
Goal: {sprint goal from _README.md}

  {icon} {id} {title:40}  {agent:15} {model:8} {notes}

Waves:
  Wave 1 {icon} ({task count})   — {verdict or "in progress" or "pending"}
  Wave 2 ...

{if learnings exists}
Learnings carry-over from previous sprints:
  - {pattern}: {suggestion}
  - ...
{endif}

{if any task blocked}
Blocked:
  {id} {title} — waiting on {dependency}
{endif}

Next action: {suggestion based on state}

Icons

• ✓ green — done
• ⋯ yellow — in progress
• ○ gray — todo
• ⊘ red — blocked
• ✗ red — failed / REQUEST CHANGES