About plan-eng-review
plan-eng-review is a skill for reviewing engineering plans against existing implementation and relevant checks. It helps teams make plans executable while preserving scope and settled decisions.
Key features
- 01Checks plans against existing code
- 02Reviews relevant engineering checks
- 03Identifies concrete gaps and corrections
- 04Preserves scope and prior decisions
- 05Defines acceptance criteria and verdicts
Use cases
- 01Assessing a plan before implementation
- 02Finding missing failure and test criteria
- 03Reviewing plans without changing source code
Install
Copy the command and run it in a terminal.
npx skills add https:/ / github.com/ borkweb/ skills --skill plan-eng-reviewAfter installing, type/plan-eng-reviewin Claude Code to confirm it loaded.
Bundled fileschecks.md
Details
From the project READMEEngineering Plan Review
Make the implementation plan coherent, testable and ready to execute. Preserve the complete requested scope; propose reductions with reasons rather than silently dropping requirements.
Scope and pacing
Read the supplied plan and relevant existing implementation. For a plan-only request, do not require a branch. For a hybrid request, pin the requested diff as well. Preserve settled constraints, protected paths and prior decisions.
Use the user's requested interaction mode. Otherwise batch independent findings and ask only about genuine unresolved choices that affect scope or design. Group up to four independent questions into one question prompt rather than one prompt each, and fold TODO or deferral rulings into the final checkpoint; batching keeps every question. Apply obvious plan corrections when improvement is requested; report-only means no edits. Honor “skip questions” immediately, state assumptions and continue useful work. An assumption does not authorize a scope change. Do not re-ask a decision already made.
Review plans, not implementation: do not modify source, commit, push or create remote tasks. Stop after the relevant acceptance criteria are addressed or report the remaining blocker. Ratings are optional summaries, not targets that require repeated polishing to 10/10.
Review
Identify the minimal working set and existing code that solves part of the problem. Verify important framework/API assumptions against installed code or authoritative documentation. File or class counts alone are not evidence of overengineering.
Read the relevant sections of checks.md: architecture/contracts, security, quality/tests, failure/concurrency, performance/operations and deployment/migration. Full mode covers every applicable section; quick mode prioritizes changed boundaries, failure paths and acceptance tests. State any omitted material coverage. Do not force all sections onto a small prose or configuration change.
For each real gap, identify the concrete trigger, file/plan reference, consequence and correction. For genuine choices, give viable options and a recommendation. A scope ruling remains settled throughout the review.
Deliver
Return or update the plan with the chosen approach, reusable code, relevant failure/verification table, dependencies, acceptance criteria and unresolved decisions. Record deferred work only when actually considered; do not invent a backlog. Add TODOs only within the requested document scope or existing authorization.
Verdict: READY when the required decisions and execution criteria are settled; NEEDS DECISIONS for material open choices; INCOMPLETE for missing required evidence. Plan readiness is not proof that implementation or deployment has passed.