Interactive multi-session orchestration. Select issues, choose terminal count, pick mode (interactive/autonomous), then spawn workers. Use when you want guided parallel issue processing.
/plugin marketplace add GGPrompts/TabzBeads/plugin install conductor@tabz-beadsThis skill inherits all available tools. When active, it can use any tool Claude has access to.
Spawn parallel workers with interactive configuration. You select issues, terminal count, and execution mode.
/conductor:bd-conduct
1. Get ready issues → bd ready
2. Select issues → AskUserQuestion (multiSelect)
3. Choose terminals → AskUserQuestion (2-4)
4. Choose mode → AskUserQuestion (Interactive/Autonomous)
5. Create worktrees → scripts/setup-worktree.sh (parallel)
6. Spawn workers → TabzChrome API
7. Send prompts → Skill-aware prompts
8. Execute mode → Interactive: monitor | Autonomous: wave loop
9. Complete → /conductor:bdc-wave-done
bd ready --json | jq -r '.[] | "[\(.priority)] \(.id): \(.title)"'
Present three AskUserQuestion prompts in sequence:
question: "Which issues should workers tackle in parallel?"
header: "Issues"
options: [
// Build from bd ready: { label: "<id> (P<priority>)", description: "<title>" }
{ label: "PROJ-abc (P1)", description: "Fix critical bug in auth" },
{ label: "PROJ-def (P2)", description: "Add dark mode toggle" }
]
question: "How many worker terminals should I spawn?"
header: "Terminals"
options: [
{ label: "2 terminals", description: "Conservative - less resource usage" },
{ label: "3 terminals (Recommended)", description: "Balanced performance" },
{ label: "4 terminals", description: "Maximum parallelism" }
]
question: "How should I run the workers?"
header: "Mode"
options: [
{ label: "Interactive (Recommended)", description: "You monitor, can intervene" },
{ label: "Autonomous", description: "Runs waves until backlog empty" }
]
After user configuration:
CONDUCTOR_SESSION=$(tmux display-message -p '#{session_name}')
# Verify skills before prompting
${CLAUDE_PLUGIN_ROOT}/scripts/match-skills.sh --available-full
# Create worktrees (parallel)
for ISSUE_ID in $SELECTED_ISSUES; do
${CLAUDE_PLUGIN_ROOT}/scripts/setup-worktree.sh "$ISSUE_ID" &
done
wait
Spawn workers via TabzChrome API:
TOKEN=$(cat /tmp/tabz-auth-token)
WORKTREE_PATH="$(pwd)-worktrees/$ISSUE_ID"
RESPONSE=$(curl -s -X POST http://localhost:8129/api/spawn \
-H "Content-Type: application/json" \
-H "X-Auth-Token: $TOKEN" \
-d "{\"name\": \"worker-$ISSUE_ID\", \"workingDir\": \"$WORKTREE_PATH\", \"command\": \"BD_SOCKET=/tmp/bd-worker-$ISSUE_ID.sock CONDUCTOR_SESSION='$CONDUCTOR_SESSION' claude --dangerously-skip-permissions\"}")
SESSION_NAME=$(echo "$RESPONSE" | jq -r '.terminal.ptyInfo.tmuxSession // .terminal.id')
# Record in beads
bd update "$ISSUE_ID" --status in_progress
bd update "$ISSUE_ID" --notes "conductor_session: $CONDUCTOR_SESSION
worker_session: $SESSION_NAME
started_at: $(date -Iseconds)"
Send skill-aware prompt (after 4s init):
sleep 4
tmux send-keys -t "$SESSION_NAME" -l "$WORKER_PROMPT"
sleep 0.3
tmux send-keys -t "$SESSION_NAME" C-m
Worker prompt structure - see references/worker-architecture.md.
${CLAUDE_PLUGIN_ROOT}/scripts/monitor-workers.sh --summarybd show $ISSUE_ID --json | jq -r '.[0].status'bd ready returns emptyWhen all selected issues are closed:
/conductor:bdc-wave-done $SELECTED_ISSUES
This runs: verify → kill sessions → merge → build → review → cleanup → push → summary
| Aspect | bd-conduct | bd-swarm-auto |
|---|---|---|
| Issue selection | User picks (multiSelect) | All ready issues |
| Terminal count | User chooses (2-4) | Auto-calculated |
| Mode | User chooses | Always autonomous |
| Entry point | Interactive prompts first | Immediate execution |
| Command | Purpose |
|---|---|
/conductor:bd-swarm-auto | Fully autonomous, no prompts |
/conductor:work | Single-session, you do the work |
/conductor:bd-work | Spawn one visible worker |
/conductor:bdc-wave-done | Complete a wave of workers |
Execute this workflow now. Start by getting ready issues.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.