メニュー
Structured planning mode for autonomous task execution. Creates plans as MemoryDocs, executes via Missions, tracks progress with live checklist.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
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。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/nearai/ironclaw --skill plan-modeコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
SOC 職業分類に基づく
SKILL.md を表示中
| name | plan-mode |
| version | 0.1.0 |
| description | Structured planning mode for autonomous task execution. Creates plans as MemoryDocs, executes via Missions, tracks progress with live checklist. |
| activation | {"keywords":["[PLAN MODE]","plan mode","create a plan","make a plan","execution plan","step by step plan"],"patterns":["\\[PLAN MODE\\]","plan (out|how to|before|for)"],"tags":["planning","autonomous","task-management"],"max_context_tokens":2500} |
You are in plan mode. Follow these protocols precisely.
When asked to create a plan (message contains [PLAN MODE] Create):
memory_search for relevant prior work and decisions.memory_write to save the plan at plans/<slug>.md where slug is a short kebab-case name derived from the goal.plan_update with status "draft" and all steps as "pending"./plan approve to start autonomous execution, or /plan revise <slug> <feedback> to adjust."Write plans to workspace memory in this format:
plan_id: <slug>
status: draft
## Goal
<clear statement of what needs to be accomplished>
## Success Criteria
<how to know the plan is complete>
## Steps
1. [ ] Step title -- tools: [tool1, tool2] -- risk: low -- est: 5min
2. [ ] Step title -- tools: [tool3] -- risk: medium -- est: 10min
3. [ ] Step title -- tools: [tool4, tool5] -- risk: low -- est: 5min
## Risks
- Risk description and mitigation strategy
## Progress Log
(updated during execution)
When asked to approve a plan (message contains [PLAN MODE] Approve):
memory_search or memory_read.mission_create with:
plan:<slug>manualmission_fire with the mission ID to start execution.plan_update with status "executing" and the mission_id.memory_write./plan status <slug>."When you are executing as part of a mission thread (your context includes "# Mission:" header with a plan):
current_focus -- if set, this tells you which step to work on next.plan_update to update the checklist:
plan_update marking the step as "failed" with the errorplan_update with overall status "failed" and stopWhen asked for plan status (message contains [PLAN MODE] Show status):
memory_search for the plan slug or "plan:".mission_list to check mission state.plan_update to refresh the UI checklist.When asked to list plans (message contains [PLAN MODE] List all plans):
memory_search with query "plan" to find plan documents./plan <description> to create one."When asked to revise (message contains [PLAN MODE] Revise):
memory_write (append: false).plan_update with status "draft" and updated steps./plan approve to re-execute.