launch-super-planner
Launch an autonomous planning agent that designs architecture and creates implementation plans as .md files for non-trivial tasks.
Instructions
Launch an autonomous planning agent. Designs architecture and creates implementation plans as .md files. Always uses claude-opus-4.6 regardless of model parameter. Use this for any non-trivial task — if the work touches 3+ files or has ambiguous requirements, plan first.
Workflow: researcher → PLANNER → coder → tester
Output goes to .agent-workspace/plans/[topic]/. After completion, read task:///{id} to get workspace path, then pass ALL .md files from that workspace as context_files to launch-super-coder. Send everything — builder-briefing.md, tester-checklist.md, task specs, the full workspace.
Status: Read task:///all every ~30s to monitor all tasks (status, deps, questions). Statuses: running, waiting → <dep>, waiting_answer ⏸, completed, failed.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| cwd | No | Working directory override (absolute path). Usually omit — server auto-detects project root. Set only if the agent needs a different root. | |
| model | No | Ignored — planner always uses claude-opus-4.6. Parameter kept for backward compatibility only. | |
| labels | No | Tags for grouping related tasks, e.g. "auth", "frontend", "v2-migration". Max 10 labels, 50 chars each. | |
| prompt | Yes | Planning brief. MUST include: PROBLEM STATEMENT (what to solve), CONSTRAINTS (what's ruled out), VERIFIED FACTS (known info), SCOPE (in/out), EXPECTED OUTPUT (what coder needs). Min 300 chars. | |
| depends_on | No | Task IDs that must complete before this starts. Handles execution ORDER only — you still must specify context_files with known paths or spawn after reading predecessor output via task:///{id}. | |
| context_files | No | Optional reference files (research docs, existing specs). Pass ALL files from prior researcher workspace — don't filter. Max 20 files, 200KB each, 500KB total. |