Agent skill

bmad:sprint

Autonomous story execution with Codex validation. Runs create/implement/review cycles per story in a bash loop until done or halted.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/bmad-sprint

SKILL.md

BMAD Sprint

Run stories autonomously with Codex validation at key boundaries. The loop is bash, state lives in files, Claude gets fresh context per session.

Usage

bash
# Run all stories for an epic, AFK mode
/bmad:sprint --epic 1 --yolo

# Run a single story interactively
/bmad:sprint --story 1-1-init-project

# Skip Codex validation (faster, less safe)
/bmad:sprint --epic 1 --skip-codex

# Resume interrupted sprint
/bmad:sprint --resume

# Full run, all epics, overnight
/bmad:sprint --yolo

What It Does

For each story in the queue:

  1. Session 1: Create + Validate - Creates story from epic, validates quality (max 3 fix cycles)
  2. Codex Boundary 1 - Reviews story spec for completeness
  3. Session 2: Test-Plan + Implement - ATDD RED phase (failing tests), then GREEN phase (make pass)
  4. Codex Boundary 2 - Reviews code for bugs/security
  5. Session 3: Test-Review + Code-Review - Quality scoring + senior review
  6. Commit - Git commit per story

When all stories in an epic complete:

  • Epic Completion: debt-fixer, test-trace, retrospective, docs-updater

Halt Conditions

Always halts (even in yolo):

  • Codex [HALT] after max fix rounds
  • Unrecoverable errors
  • Blocked stories (needs human)
  • No signal from Claude

Mode-dependent:

  • Interactive: pauses between stories for review
  • Yolo: auto-continues through everything

State Tracking

Sprint progress tracked in _bmad-output/implementation-artifacts/tracking/SPRINT.md:

  • Current story and session
  • Story completion status
  • Codex validation results
  • Checkpoints for resume

Story statuses in _bmad-output/implementation-artifacts/tracking/sprint-status.yaml.

Prerequisite

Before first run, generate sprint-status.yaml:

/bmad-bmm-sprint-planning

Execution

This runs in a tmux session so you can detach and reattach.

bash
ARGS="$ARGUMENTS"
SESSION_NAME="bmad-sprint"

# Find the script (project-local or global)
if [[ -f ".claude/skills/bmad-sprint/scripts/sprint.sh" ]]; then
  SCRIPT=".claude/skills/bmad-sprint/scripts/sprint.sh"
else
  SCRIPT="$HOME/.claude/skills/bmad-sprint/scripts/sprint.sh"
fi

# Kill existing session if running, create new one
tmux kill-session -t "$SESSION_NAME" 2>/dev/null || true
tmux new-session -d -s "$SESSION_NAME" -c "$(pwd)"

# Run the sprint script in the tmux session
tmux send-keys -t "$SESSION_NAME" "bash $SCRIPT $ARGS" Enter

# Tell user how to attach
echo ""
echo "================================================================"
echo "BMAD Sprint started in tmux session: $SESSION_NAME"
echo "================================================================"
echo ""
echo "To attach and watch progress:"
echo "  tmux attach -t $SESSION_NAME"
echo ""
echo "To detach (while attached):"
echo "  Ctrl+b then d"
echo ""
echo "To check if still running:"
echo "  tmux has-session -t $SESSION_NAME 2>/dev/null && echo 'Running' || echo 'Finished'"
echo ""

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results