菜单
Pre-flight checklist and post-implementation self-review protocol. Use before generating any code (pre-flight) and after writing code but before verification (self-review) to catch issues early.
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。
Structured fault tolerance for coordinator agents. 5-level escalation ladder (Retry → Replace → Skip → Redistribute → Degrade), dead-man timers, degraded completion protocol, and cross-level escalation format. Load when orchestrating agents that may fail.
Structured code review protocol for inspecting code quality against the full rule set. Use when auditing code written by yourself or another agent, during the /audit workflow, or when the user asks for a code review.
Reusable convergence protocol for coordinator agents. Defines the BRIEFING → ITERATE → GATE → CONVERGE loop, context hygiene rules, self-succession protocol, turn budget, and handoff compression. Load when orchestrating multi-iteration workflows.
MECE task decomposition, file ownership enforcement, DAG-based execution, and safe merge protocol for intra-domain parallel dispatch. The safety invariants that prevent merge chaos when multiple agents write in parallel. Applies recursively at every nesting depth.
Shared protocols for all agents in the multi-agent pipeline: recursive nesting, pre-implementation restatement, parallel dispatch format, and agent definition cascade. Load this skill instead of inlining these protocols in every agent file.
Profile-driven performance optimization protocol. Use when profiling data (CPU, heap, trace) is available or when the user requests performance analysis. Covers methodology (Profile → Analyze → Opportunity Scan → Prioritize → Optimize → Benchmark), optimization pattern catalog, safety invariants, and when-to-stop heuristics. Language-specific tooling is in languages/*.md.
| name | guardrails |
| description | Pre-flight checklist and post-implementation self-review protocol. Use before generating any code (pre-flight) and after writing code but before verification (self-review) to catch issues early. |
Structured checklists that ensure the agent considers all applicable rules before and after writing code. Catches issues that would otherwise only surface during verification.
Run through this checklist before writing any code:
.agents/rules/architectural-pattern.md)project-structure.md)rule-priority.md for any potential conflictsIf any item cannot be checked, stop and resolve before proceeding.
Run through this checklist after writing code, before verification:
For universal error handling principles, see
.agents/rules/error-handling-principles.md. Below: language-specific patterns only.
project-structure.mdAfter completing the universal checklist above, load the relevant language-specific checklist from code-review/languages/{language}.md. The language files contain both review anti-patterns and self-review checklists.
Only load the file for languages you are actively writing. If a language is not listed, skip — but the universal checklist above always applies.
This skill enforces: