跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
搜索与研究

plan-task

Plans a task or feature - loads context, optionally clarifies requirements, spawns Plan agent, persists plan to .groundwork-plans/

etr · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add etr/groundwork --skill plan-task

技能说明


name: plan-task description: Plans a task or feature - loads context, optionally clarifies requirements, spawns Plan agent, persists plan to .groundwork-plans/ requires: understanding-feature-requests argument-hint: "[task-number-or-description] [--project name]"

Task Planning Skill

Plans a task or feature by loading context, optionally clarifying requirements, spawning a Plan agent, and persisting the validated plan to .groundwork-plans/.

Explicit Project Input

If arguments include --project <name>, resolve that project directly from the repository's .groundwork.yml for this invocation. Treat it as authoritative; do not depend on or change persisted project selection. In runner mode, GROUNDWORK_PROJECT and GROUNDWORK_PROJECT_ROOT provide the same invocation-local selection.

Runner Mode

If session context contains GROUNDWORK_RUNNER_MODE=true:

  • Skip the model recommendation pre-flight.
  • Treat GROUNDWORK_BATCH_MODE=true as authoritative: do not call AskUserQuestion.
  • When information or a decision is missing, return RESULT: FAILURE instead of asking.
  • Do not modify .gitignore; the runner keeps .groundwork-plans/ out of Git state.
  • Keep the existing terminal RESULT: PLANNED contract unchanged.

Token Discipline

This skill orchestrates planning workflows. Every turn re-reads the full context window, so unnecessary turns are expensive.

  1. No narration turns. Do not output text-only turns like "Let me load the task" or "Now I'll spawn the plan agent." Combine text with a tool call in the same turn, or skip the text entirely.
  2. Batch tool calls. When multiple tool calls are independent, issue them all in one turn.
  3. No waiting updates. Do not output "Waiting for results..." turns. Wait silently until results arrive.
  4. Keep context lean. Do not read file contents you won't use directly. Pass file paths to subagents and let them read in their own context windows.

Pre-flight: Model Recommendation

Your current effort level is {{effort_level}}.

Skip this step silently if effort is high, xhigh, or max (the scale is low < medium < high < xhigh < max, so xhigh and max are already above high) AND you are Sonnet or Opus. If effort is low or medium (i.e. below high), you MUST show the recommendation prompt — regardless of model. If you are not Sonnet or Opus, you MUST show the recommendation prompt — regardless of effort level.

Otherwise → use AskUserQuestion:

{
  "questions": [{
    "question": "Planning benefits from consistent multi-domain reasoning.\n\nTo switch: cancel, run `/effort high` (and `/model sonnet` if on Haiku), then re-invoke.",
    "header": "Effort check",
    "options": [
      { "label": "Continue" },
      { "label": "Cancel — I'll switch first" }
    ],
    "multiSelect": false
  }]
}

If the user selects "Cancel — I'll switch first": output the switching commands and stop. Do not proceed with the skill.

Step 0: Resolve Project Context

Before loading specs, ensure project context is resolved:

  1. Monorepo check: Does .groundwork.yml exist at the repo root?
    • If yes → Is {{project_name}} non-empty?
      • If empty → Invoke Skill(skill="groundwork:select-project") to select a project, then restart this skill.
      • If set → Project is {{project_name}}, specs at {{specs_dir}}/.
    • If no → Continue (single-project repo).
  2. CWD mismatch check (monorepo only):
    • Skip if not in monorepo mode or if the project was just selected in item 1 above.
    • If CWD is the repo root → fine, proceed.
    • Check which project's path CWD falls inside (compare against all projects in .groundwork.yml).
    • If CWD is inside the selected project's path → fine, proceed.
    • If CWD is inside a different project's path → warn via AskUserQuestion:

      "You're working from <cwd> (inside [cwd-project]), but the selected Groundwork project is [selected-project] ([selected-project-path]/). What would you like to do?"

      • "Switch to [cwd-project]"
      • "Stay with [selected-project]" If the user switches, invoke Skill(skill="groundwork:select-project").
    • If CWD doesn't match any project → proceed without warning (shared directory).
  3. Proceed with the resolved project context. All {{specs_dir}}/ paths will resolve to the correct location.

Step 1: Detect Input Mode

Parse the input from the caller's conversation context. Three modes:

  • Task ID (numeric like 4, or full format like TASK-004): Set mode=task, task_id=TASK-NNN (zero-padded to 3 digits).
  • Blurb (free text, no task ID): Set mode=feature, blurb=<text>.
  • Task ID + blurb (both present): Set mode=task+blurb, task_id=TASK-NNN, blurb=<text>.
  • Neither (no argument, no context): Output RESULT: FAILURE | No task ID or feature description provided. and stop.

Step 2: Load Context (mode-dependent)

Task ID mode (mode=task or mode=task+blurb)

  1. Locate task file:

    • Single file: {{specs_dir}}/tasks.md
    • Directory: {{specs_dir}}/tasks/ Search for ### TASK-NNN: pattern. Error: Task not found → Output RESULT: FAILURE | TASK-NNN not found in {{specs_dir}}/tasks/ and stop.
  2. Read task definition — only the ### TASK-NNN: section (goal, action items, acceptance criteria, dependencies, status). Do NOT read full specs/architecture/design files.

  3. Validate task is workable:

    • Status Complete → Output RESULT: FAILURE | TASK-NNN is already Complete. and stop.
    • Blocked by incomplete tasks → Output RESULT: FAILURE | TASK-NNN is blocked by: [list]. and stop.
  4. Verify spec paths exist (use Glob, not Read):

    • {{specs_dir}}/product_specs.md or {{specs_dir}}/product_specs/
    • {{specs_dir}}/architecture.md or {{specs_dir}}/architecture/
    • {{specs_dir}}/design_system.md
    • Tasks file (already found above)

    If specs missing: Report which are m

阅读完整来源 ↗