一键导入
implement
Execute a checked-in implementation plan end-to-end using parallel agents. Use when a spec and plan exist in .plans/ and you're ready to build.
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
Execute a checked-in implementation plan end-to-end using parallel agents. Use when a spec and plan exist in .plans/ and you're ready to build.
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
Cut a PPDS release — CHANGELOG refresh, version bump, tag push sequence, CI monitoring, post-publish verification. Use when preparing a new prerelease or stable release across CLI, TUI, MCP, Extension, and NuGet libraries.
Triage and merge open dependabot PRs per docs/MERGE-POLICY.md — classify each PR (auto-merge / verify-then-merge / manual review), enable auto-merge for safe ones, run targeted test suites for risky ones, and surface anything needing human judgment. Use when there's a backlog of dependabot PRs, after a quiet period, or as a routine drain.
AI self-verification of implemented work across surfaces (extension, CLI, MCP, TUI, workflow). Use after implementation to verify code works in its runtime environment.
Create, triage, and manage GitHub issues per PPDS backlog conventions. Use when creating issues, grooming the backlog, reviewing what to work on next, or filing bugs.
Brainstorm ideas into specs and plans through collaborative dialogue. Use when starting a new feature, exploring an idea, or designing a system — before any implementation.
Gates
| name | implement |
| description | Execute a checked-in implementation plan end-to-end using parallel agents. Use when a spec and plan exist in .plans/ and you're ready to build. |
/implement — read spec and plan from .plans/, execute phases
/implement .plans/<date>-<name>.md — explicit plan path
/implement specs/<name>.md — spec path: generate the plan from this spec first, save it to .plans/, then execute
Agent tool, /review, /verify, /qa, /debug, /gates.
$ARGUMENTS = path to a plan file, or a spec file to generate the plan from. If omitted: use the most relevant spec on the branch, generate a plan, save to .plans/, proceed.
Fallback — no spec: check .plans/context.md first; if present, generate a plan from it. Otherwise: prompt the user — run /design or continue without spec? See REFERENCE.md §1 for the full fallback chain.
Read specs/CONSTITUTION.md + relevant specs (grep **Code:** frontmatter). Build spec context block for every subagent. See REFERENCE.md §4.
Check git status, branch, existing worktrees, prior phase commits.
Build task list from plan phases; mark already-completed work done.
Read REFERENCE.md §3 for Opus vs. Sonnet guidance.
A. Dispatch Agents — parallel for independent tasks; see REFERENCE.md §5.
B. Collect Results — wait for all agents; review summaries.
B2. Cross-Agent Consistency Check — verify cross-surface contract consistency; see REFERENCE.md §7.
C. Verify Phase Gate — build → tests → AC coverage → surface-specific verify/qa → review. See REFERENCE.md §6. Fix before advancing.
D. Review — invoke /review (reviewer sees diff + constitution + ACs only). Fix issues.
E. Commit — git add specific files; commit per REFERENCE.md §8.
F. Advance — move to next phase after commit. Update task tracking.
A. Gates — /gates
B. Verify — /verify extension|tui|cli|mcp per changed surfaces
C. QA — /qa extension|cli|mcp|tui per changed surfaces
D. Review — /review final comprehensive review
E. Converge — if critical/important findings: gates→review→fix loop (max 5 cycles)
F. Final State Check — git status --porcelain produces no output (no uncommitted or untracked files); all phases committed.
G. Submit — proceed IMMEDIATELY to /pr; do not stop to summarize