メニュー
Spawn a subagent to run a task. Use when you want to delegate work to a separate pi instance.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Use when opening a PR and driving it to green CI — push the branch, write a high-level PR description, watch checks, fix failures, and keep looping until green; also use when an already-open PR gets new commits.
Use when the user starts, resumes, switches, saves, or recalls a named task or ongoing work across sessions (e.g. "remember this", "continue X", "what was I doing on Y", "track this task").
Git commit workflow. Load when finishing any code-change task (after verification) to commit and push, or when explicitly staging/committing.
Use when modifying or adding a transform, loss, model, method, package, task-model, config, docs page, or example in a Lightly AG repo (lightly-ssl or lightly-train) and you need repo-specific conventions.
Use when configuring or replicating LT-DETR v2 (ltdetrv2-s/m/l/x) COCO benchmarks via the lightly-train wrapper — model alias map, global vs per-rank batch size, recipe invariants, ECDet config mapping, cluster pins, and common gotchas.
Use when the user starts, resumes, switches, saves, or recalls a named task or ongoing work across sessions (e.g. "remember this", "continue X", "what was I doing on Y", "track this task").
| name | subagent |
| description | Spawn a subagent to run a task. Use when you want to delegate work to a separate pi instance. |
Spawn a subagent to run a task independently.
The script is at ~/.pi/agent/skills/subagent/subagent.sh
./subagent.sh "your prompt here"
--model <model> - Model to use (use minimax by default, minimax/MiniMax-M3)--thinking <level> - Thinking level: off, low, medium, high--cwd <dir> - Working directory./subagent.sh "List files in src/"
./subagent.sh "Find all TODO comments" --model minimax/MiniMax-M3
For broad read-only searches (repo + git history + web), the subagent can take several minutes and block the main turn. Two patterns avoid the hang:
watch: redirect output to a temp file and launch the
subagent through the watch tool (process mode). The agent turn continues
immediately and resumes when the subagent exits:
./subagent.sh "<prompt>" --cwd <dir> --thinking high > /tmp/sub.txt 2>&1
# wrap the line above in a `watch` (process mode) call
If no output appears after a few minutes, cancel the watcher and continue with main-agent verification. Prefer narrow prompts; broad "search repo + history