agent-standup

name	agent-standup
description	Skill for running standup meetings across all active agents in the swarm. Aggregates Beads task status, git branch activity, commit history, and conflict detection into a human-readable progress report.
user-invocable	true
disable-model-invocation	false

Agent Standup

Use this skill when running a progress check or standup meeting across all active agents in the swarm.

Running a Standup

Use Beads and git directly to gather status:

# Task status
bd list --json | jq -r '.[] | "[\(.id)] \(.status) P\(.priority) \(.title)"'

# Ready tasks
bd ready

# Per-branch activity
for wt in ../$(basename $(pwd))-swarm/agent-*; do
  [ -d "$wt" ] || continue
  b=$(basename "$wt")
  echo "$b: $(git log main..$b --oneline 2>/dev/null | wc -l | tr -d ' ') commits"
done

Or run the dashboard for live monitoring:

python3 scripts/dashboard.py

Interpreting Results

Healthy Agent

• Regular commits (multiple in the last few hours)
• Beads task status matches git activity
• No conflicts with other branches

Stalled Agent

• No commits in the last 24 hours
• Task still marked as in-progress
• Action: Consider restarting with /spawn

Blocked Agent

• Beads notes contain "BLOCKED"
• No recent commits
• Action: Check dependency graph, merge blockers first

Conflict Detected

• Two agent branches modify the same files
• Action: Merge one branch first, then rebase the other

Updating Beads After Standup

bd update <id> --notes "Standup $(date +%Y-%m-%d): [status summary]"

Automated Standups

Standups run as part of the orchestrator's OODA loop — the orchestrator Copilot checks agent status continuously and acts on stalls, blocks, and completions.