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。
浏览器与自动化

executing-plans

Use when you have a written implementation plan to execute in a separate session with review checkpoints

DollarDill · 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 DollarDill/beads-superpowers --skill executing-plans

技能说明


name: executing-plans description: Use when you have a written implementation plan to execute in a separate session with review checkpoints

Executing Plans

Overview

Load plan, review critically, execute all tasks, report when complete.

Announce at start: "I'm using the executing-plans skill to implement this plan."

Note: Tell your human partner that this system works much better with access to subagents (Claude Code, Codex, Copilot CLI, and Gemini CLI all qualify). If subagents are available, use beads-superpowers:subagent-driven-development instead of this skill.

The Process

Step 1: Load and Review Plan

  1. Ensure an isolated workspace first — use the beads-superpowers:using-git-worktrees skill.

  2. Read plan file

  3. Review critically - identify any questions or concerns about the plan

  4. If concerns: Raise them with your human partner before starting

  5. If no concerns: Create epic bead and child beads for each task, then proceed

    Create the epic, then import the child tasks as JSONL:

    # 1. Create the epic (→ note its id):
    bd create "Epic: <plan-name>" -t epic -p 2 -d "<goal>
    

Success Criteria

  • <measurable outcome from the plan's Goal>"

    2. Author the tasks as JSONL — one issue per line, id OMITTED (auto-assigned;

    a supplied colliding id would overwrite that bead and reset its omitted fields).

    Parent each to the epic; embed the bd lint-required '## Acceptance Criteria' in

    'description' (or the 'acceptance_criteria' field). Read bd import --help on

    first use; bd export <id> round-trips a real bead as a schema template.

    cat <<'EOF' | bd import - {"title":"Task 1: ","issue_type":"task","priority":2,"description":"\n\n## Acceptance Criteria\n- ","dependencies":[{"depends_on_id":"","type":"parent-child"}]} {"title":"Task 2: ","issue_type":"task","priority":2,"description":"\n\n## Acceptance Criteria\n- ","dependencies":[{"depends_on_id":"","type":"parent-child"}]} EOF

    
    Confirm the import output shows no `Skipped dependency` (a dep to a missing target is skipped, not rolled back). `bd lint` requires `## Success Criteria` in the epic's description and `## Acceptance Criteria` in each task's — embed them as above.
    
    > **Wire task ordering (`blocks`) after the import.** `parent-child` rides the import, but inter-task `blocks` deps do not (children's ids are auto-assigned, so siblings can't reference each other in one file). Capture the ids **scoped to the parent** (a title grep collides with old beads), then wire ordering atomically:
    > ```bash
    > bd ready --parent <epic-id> --json   # → the child task ids
    > printf 'dep add <task-2-id> <task-1-id> blocks\n' | bd batch
    > ```
    > Note: `bd batch create` does not support `--description`/`--parent`/`--acceptance` — that is why task *creation* uses `bd import`, not `bd batch`.
    

Step 2: Execute Tasks

For each task:

  1. Get and claim the next task in one call: bd ready --parent <epic-id> --claim (use bd ready --explain to see dependency reasoning if task ordering is unclear)
  2. Check description quality before implementing: if the claimed task's description is a bare title with no actionable steps or context, STOP — do not proceed with implementation. The task is now claimed: flag it for human decision (bd label add <task-id> human, per Structured blocker handling below) so it doesn't dangle in-progress, and surface to the user what the description is missing.
  3. Follow each step exactly (plan has bite-sized steps), using the beads-superpowers:test-driven-development skill for the implementation — RED-GREEN-REFACTOR
  4. Run verifications as specified
  5. Close the task: bd close <task-id> --reason "description of what was completed"
  6. Check epic progress: bd epic status <epic-id> to see overall completion

--claim consent boundary. This skill's autonomous take-next / batch-dispatch flow is the one place bd ready --claim is legitimate. That autonomous --claim is FORBIDDEN wherever the user picks the work (orientation, brainstorming, session close) — the consent gate binds even when this skill is not loaded.

Step 3: Complete Development

After all tasks complete and verified:

  • Announce: "I'm using the finishing-a-development-branch skill to complete this work."
  • REQUIRED SUB-SKILL: Use beads-superpowers:finishing-a-development-branch
  • The finishing skill includes the Land the Plane session close protocol (bd close → bd dolt push → git push → git status)
  • Follow that skill to verify tests, present options, execute choice

When to Stop and Ask for Help

STOP executing immediately when:

  • Hit a blocker (missing dependency, test fails, instruction unclear)
  • Plan has critical gaps preventing starting
  • You don't understand an instruction
  • Verification fails repeatedly

Structured blocker handling: When you hit a blocker, classify it and use the appropriate response:

Blocker typeActionCommand
Time-based (waiting on deploy, external process)Defer the task for laterbd defer <task-id> --until="<date>"
Missing work (prerequisite not built yet)Create the missing task and wire dependencybd create "Missing: <title>" -t task --parent <epic-id> then bd dep add <blocked-id> <new-id>
Human decision needed (architecture choice, ambiguous requirement)Flag for human inputbd label add <task-id> human

Discovered-work bead stamp: bd create "[spec] <title>" -t task --parent <epic-id> --notes "Severity:/Confidence:/Evidence:" — see verification-before-completion → Agent-Filed Bead Discipline.

Ask for clarification rather than guessing.

When to Revisit Earlier Steps

Return to Review (Step 1) when:

  • Partner updates the plan based on your feedback
  • Fundamental approach needs rethinking

Don't force through blockers - stop a

阅读完整来源 ↗