一键导入
executing-plans
This skill should be used when the user has a written implementation plan to execute in a separate session with review checkpoints.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
This skill should be used when the user has a written implementation plan to execute in a separate session with review checkpoints.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
This skill should be used when the user needs deep, multi-step web research with source synthesis, citations, skeptical evidence evaluation, confidence/gap analysis, and optional dense/frontier research using parallel agents.
This skill should be used when the user wants to see all installed plugins, agents, skills, and MCP servers, and also inspect the current repository for local agents, skills, and MCP configuration. Scans the environment and presents a structured catalog of available resources. Works across all AI CLI platforms (Claude Code, GitHub Copilot, Gemini CLI, OpenCode, OpenAI Codex).
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
This skill should be used when the user needs to transform audio recordings into professional Markdown documentation with intelligent summaries using LLM integration.
Relentless Socratic interview that walks down every branch of a plan or design, resolving decisions one-by-one. Trigger ONLY when the user explicitly says "grill me" or the PT equivalent "me grille". Do NOT trigger for stress-test, poke holes, devil's advocate, challenge, review, or feedback requests.
This skill should be used when the user needs structured strategic analysis and high-impact executive recommendations for complex business problems.
| name | executing-plans |
| description | This skill should be used when the user has a written implementation plan to execute in a separate session with review checkpoints. |
| license | MIT |
Execute approved implementation plans in controlled batches with checkpoints, verification, and review gates.
Use this skill when:
Display plan execution progress before each batch:
[████░░░░░░░░░░░░░░░░] 20% — Phase 1: Loading & Reviewing Plan
[████████░░░░░░░░░░░░] 40% — Phase 2: Executing Batch N
[████████████░░░░░░░░] 60% — Phase 3: Reporting & Verification
[████████████████░░░░] 80% — Phase 4: Awaiting Feedback
[████████████████████] 100% — Phase 5: Final Verification Complete
For plans with many tasks, adapt the gauge to reflect actual task completion: X/Y tasks done (N%).
Load plan, review critically, execute tasks in batches, report for review between batches.
Core principle: Batch execution with checkpoints for architect review.
Announce at start: "I'm using the executing-plans skill to implement this plan."
Default: First 3 tasks
For each task:
When batch complete:
Based on feedback:
After all tasks complete and verified:
STOP executing immediately when:
Ask for clarification rather than guessing.
Return to Review (Step 1) when:
Don't force through blockers - stop and ask.
Required workflow skills:
| Error | Likely Cause | Action |
|---|---|---|
| Plan file not found | Path provided doesn't exist or file was moved | Ask user to confirm file path; list available plan files in docs/plans/ |
| Ambiguous task description | Step in plan is unclear or underspecified | Stop and ask user for clarification before proceeding |
| Task depends on failed prerequisite | Earlier task failed or was skipped | Report the dependency failure; do not continue until resolved |
| Verification command fails | Tests or checks defined in plan are failing | Stop immediately, report full error output, wait for user instruction |
| Blocker encountered mid-execution | Unexpected state, conflicting files, or missing context | Pause, report the exact blocker, and ask user how to proceed |
| Batch completed with partial failures | Some tasks in batch failed | Report which succeeded and which failed; ask user whether to retry or skip |
docs/plans/2026-02-20-api-hardening.md."