Basic Spawning
Spawn a single agent for an issue:- Creates an isolated git worktree from your default branch
- Creates a feature branch (e.g.,
feat/123) - Starts a tmux session for the agent
- Launches the AI agent with issue context from your tracker
- Writes session metadata for tracking
The issue identifier format depends on your tracker:
- GitHub:
#123or just123 - Linear:
INT-1234(team prefix + number) - Ad-hoc: Any string becomes the branch name
Spawning with Issue IDs
GitHub Issues
gh) and uses the title and description as context.
Linear Issues
Configure Linear in your project:Ad-hoc Tasks (No Issue)
Spawn without an issue ID for exploratory work:Batch Spawning
Spawn multiple agents in parallel with duplicate detection:- Duplicate detection: Skips issues that already have active sessions
- Same-batch deduplication: Won’t spawn duplicates within the batch
- Parallel execution: Creates sessions with 500ms delay between each
- Summary report: Shows created, skipped, and failed sessions
Agent Selection
Override the default agent per-session:claude-code(default) — Anthropic Claude with Code modecodex— OpenAI Codexaider— Aider coding assistantopencode— OpenCode agent
- Claude Code
- Codex
- Aider
Custom Prompts
Using Ad-hoc Sessions
For work without a tracked issue:Agent Rules (Per-Project)
Define rules that apply to all agents in a project:.agent-rules.md:
Opening Sessions in Terminal
Automatically open spawned sessions in terminal tabs:iterm2) to create a new tab and attach to the tmux session.
Manual attachment:
Session Output
When a session is created, you’ll see:SESSION= line is for scripting:
Pre-flight Checks
Before spawning, the CLI validates:
Common errors:
Advanced: Session Prefixes
Control session naming with prefixes:fe-1, api-1, mob-1
This helps when managing multiple projects:
Next Steps
Managing Sessions
List, attach, send messages, and kill sessions
Auto-Reactions
Configure automatic CI failure and review handling
