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

plan-review-gate

Use after saving a non-trivial plan or decision RFC when a fail-closed feasibility, completeness, and alignment review must block execution.

romiluz13 · 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 romiluz13/cc10x --skill plan-review-gate

技能说明


name: plan-review-gate description: "Use after saving a non-trivial plan or decision RFC when a fail-closed feasibility, completeness, and alignment review must block execution." allowed-tools: Read Bash Grep Glob user-invocable: false

Spec Review Gate

Core principle: No execution plan or decision RFC reaches the user without surviving adversarial scrutiny. No leniency. "Close enough" is FAIL. A structurally neat but repo-wrong plan is FAIL.

How it works: This skill runs inline in the calling agent's context (no subagents). The calling LLM acts as an independent auditor, reads the saved artifact, and checks it against 3 criteria using Read/Grep/Glob. The value is fail-closed blocking and adversarial framing, not reviewer isolation — keep the auditor posture, but do not claim independence the gate doesn't have.

When to Skip

Skip when artifact is truly trivial: Single-file fix, copy edit, or config tweak with <3 changes and no architecture choice → return SPEC_GATE_PASS immediately.

The 3 Checks (Run in Sequence)

Check 1: Feasibility — Can this be executed against the real codebase?

Run these verifications using Read/Grep/Glob:

CriterionHow to verifyBlocking if
Artifact file exists on diskGlob(pattern="{plan_file_path}") where plan_file_path is the path from the calling agent's contextReturns 0 matches
File paths existGlob(pattern="{path}") for every referenced fileAny path returns 0 matches and doesn't exist
Codebase reality check is presentRead artifact for Codebase Reality CheckNon-trivial plan omits repo-grounded verification of existing code
Dependency orderingRead plan phases — do later phases depend on earlier ones only?Circular or forward-reference dependencies
Technical approach matches codebaseRead 1-2 existing files in the affected areaProposed patterns/libs differ from what codebase actually uses
No unstated infra assumptionsRead plan for external services, env vars, DBsPlan silently assumes infra that doesn't exist
No invented or unverified file/module assumptionsCompare claimed touched surfaces to repo realityArtifact presents guessed files/modules as verified facts
Plan mode fits the taskRead request + artifact: direct, execution_plan, or decision_rfcRequest changes ≥3 files or any contract/schema/auth surface while mode is direct; or request asks for a decision between alternatives and mode is not decision_rfc
Verification rigor fits riskCheck verification_rigor against the requested workCritical-path work is missing critical_path rigor or claims proof it never defined

Critical-path work = auth, payment, data-destructive operations, migrations, or work the user labeled critical.

Check 2: Completeness — Does it cover the full request?

Read the user's original request and compare against the plan:

CriterionHow to verifyBlocking if
All requirements mappedList each sentence of the user request; cite the plan item covering itAny user requirement has no corresponding plan item
Verification steps definedFor each plan item, find its named test, command, or checklistAny change has no way to verify it worked
Edge cases addressedFor each input surface the plan touches, find its empty, invalid, and failure caseAn input surface's empty, invalid, or failure case is neither named nor covered by an explicit "why none applies" statement
Cross-file integrationTrace each touched surface to the callers/importers the plan namesmissing touched surfaces or integration points
Plan-vs-code gaps surfacedRead the artifact for the mismatch table; spot-check one claim against the codeA non-trivial plan omits the concrete mismatch table or hides contradictions with current code
Assumption ledger is honestRead each important claim for its classification tagImportant claims are not classified as proven_by_code, inferred, or needs_user_confirmation
Phase dependency map is presentRead each phase for its depends-on/enables statementsNon-trivial phases do not say what they depend on or what they enable
Durable Decisions present for multi-phase plansFor multi-phase plans, read for a foundational-decisions sectionA multi-phase plan omits foundational decisions (routes, schema, models, auth, third-party boundaries) that all phases should reference
Decision-grade content present when neededFor decision_rfc, read for alternatives, drawbacks, and references sectionsA decision_rfc is missing alternatives, drawbacks, or references
Critical-path spec present when neededFor critical_path, read for each of the five required sectionsA critical_path artifact is missing behavior contract, edge-case catalog, provable properties, purity boundary, or verification strategy

Check 3: Scope & Alignment — Is it right-sized and faithful?

CriterionHow to verifyBlocking if
Matches user requestRe-read the request; diff its ask against the plan's stated goalPlan solves different problem or adds unrequested features
No scope creepFor each plan item, name the requirement it servesExtra abstractions, refactoring, or features beyond the request
No under-scopingList the request's direct implications; find each in the planDirect implications of the request are omitted
Execution order is realWalk the phases in order; check each prerequisite exists in an earlier phasewrong execution order or missing prerequisites
Complexity proportionalFor each new file, abstraction, or dependency, cite the requirement row that needs itThe plan introduces a file, abstraction, or dependency that no requirement row maps to
Defaults are framed honestlyCheck each recommended
阅读完整来源 ↗