菜单
[Command entry-point - invokes shared `autopilot-pipeline` skill] Direct command-bar dispatch only; do not select for general queries.
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
Performs non-destructive cross-artifact consistency and quality analysis across spec, plan, and tasks. Also supports remediation mode to apply fixes. Use when running /sddp-analyze or when compliance auditing is needed.
Runs the full feature-delivery SDD pipeline end-to-end without user interaction. When called without arguments, auto-selects the first unchecked epic from specs/project-plan.md. Requires Autopilot enabled in config, a Product Document, and a Technical Context Document. Use when running /sddp-autopilot.
Clarify product, technical, and operational specs with targeted questions and write accepted answers into spec.md.
[DEPRECATED — rules moved to AGENTS.md §Communication Style. This file is a deprecation shim only. Do not re-introduce a Read instruction for this path; the rules are ambient context loaded once via AGENTS.md.]
Generates requirements quality checklists ('Unit Tests for English') that validate quality, clarity, and completeness in a given domain. Use when running /sddp-checklist or when quality verification of requirements is needed.
Orchestrates decomposition of implementation plans into actionable, dependency-ordered task lists. Use when running /sddp-tasks or when task generation from a plan is needed.
| name | sddp-autopilot |
| description | [Command entry-point - invokes shared `autopilot-pipeline` skill] Direct command-bar dispatch only; do not select for general queries. |
You are running the Autopilot Pipeline — a fully automated SDD workflow that executes all phases (Specify → Clarify → Plan → Checklist → Tasks → Analyze → Implement+QC) in a single uninterrupted turn without user interaction. Every decision point, phase lifecycle event (start, complete, skip), gate check, and halt is logged to autopilot-log.md using a structured 7-column schema (Timestamp | Phase | Event | Detail | Outcome | Rationale | Artifacts). Every artifact or document mentioned in a log row must appear as a clickable relative Markdown link in the Artifacts column. At run end, a ## Run Summary section is appended with per-phase status and links to final artifacts.
Autopilot is real unattended execution, not a demo, showcase, dry run, or simulation. Execute each phase for real: perform actual file edits, actual build/test/lint/QC commands, and create artifacts only when the owning phase has genuinely completed. Never simulate implementation, QC, test results, or marker creation. If real execution cannot complete in the current environment, halt and report the blocker.
Load and follow the workflow in .github/skills/autopilot-pipeline/SKILL.md.
The pipeline skill will instruct you to load and execute these sub-skills inline, in order:
.github/skills/specify-feature/SKILL.md.github/skills/clarify-spec/SKILL.md.github/skills/plan-feature/SKILL.md.github/skills/generate-checklist/SKILL.md (looped until queue exhausted).github/skills/generate-tasks/SKILL.md.github/skills/analyze-compliance/SKILL.md.github/skills/implement-qc-loop/SKILL.mdWhen any sub-skill says Delegate, read the referenced sub-agent file at that point, not before — then perform the task yourself:
.github/agents/_context-gatherer.md.github/agents/_task-tracker.md.github/agents/_developer.md.github/agents/_checklist-reader.md.github/agents/_test-evaluator.md.github/agents/_technical-researcher.md.github/agents/_qc-auditor.md.github/agents/_story-verifier.md.github/agents/_policy-auditor.md.github/agents/_test-planner.mdAUTOPILOT = true for all phases. At every user interaction point, choose the recommended default and log the decision — never prompt the user.
Report progress at each phase boundary. Only halt for the conditions defined in the pipeline skill.