code-review

Runs structured Spec Kit reviews of documentation, code, and tests against constitution rules.

1|Updated Jan 4, 2026
One-click install
npx skills add https://github.com/stefaniuk/loadout --skill code-review-stefaniuk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/stefaniuk/loadout/tree/main/.github/skills/code-review
Command: npx skills add https://github.com/stefaniuk/loadout --skill code-review-stefaniuk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Spec-driven development pipelines drift when code, specifications, plans, and tests are reviewed ad hoc. This Skill enforces a consistent, evidence-based review process that checks every artefact against the constitution and spec set, so discrepancies are caught with concrete file-and-line evidence instead of subjective opinion. ## Core Features & Use Cases - Three review modes: Run a documentation review (Product Owner role, pre-implementation), a code review (Implementation Engineer role, post-implementation), or a test review (test pyramid and coverage quality), each with its own companion workflow document. - Constitution-first analysis: Every finding cites the constitution rule or spec requirement identifier it violates, plus a workspace-relative path and line range. - Structured reports with decision checklists: Each mode produces a fixed output structure including severity-ranked findings, Proposed Resolutions (Options A/B/C), and an explicit approval checklist. - Use Case: After implementing a feature from specs/004-create-order/, run the code-type review to detect an idempotency behaviour present in code but missing from the spec, then decide via the checklist whether to align code or amend the spec. ## Quick Start Ask the assistant to run the code-review skill with type code, documentation, or test for the current feature spec folder.

Frequently Asked Questions about code-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a spec-driven code review with this skill?▼

Invoke the skill with a review type argument: code, documentation, or test. The skill reads the constitution, then the feature's spec.md, plan.md, and tasks.md in order, and produces a structured report with evidence-linked findings and a decision checklist.

What is the difference between the documentation, code, and test review types?▼

Documentation review checks spec integrity, terminology, and traceability before implementation. Code review checks constitution compliance and spec coverage after implementation. Test review assesses test pyramid health and unit test quality after the code review passes.

In what order should the review types run in the Spec Kit pipeline?▼

Run documentation review after speckit-plan and speckit-tasks but before implementation, then code review after implementation, then test review last. Each review type is a separate invocation; modes are never bundled into one run.

Does the code review modify code automatically?▼

No. The review only reports findings and proposes resolutions as Options A (align code to spec), B (align spec to code), or C (request missing evidence). Code changes require explicit user approval through the decision checklist.

What happens when information is missing during a review?▼

The skill marks gaps explicitly as "Unknown from code - {suggested action}" instead of guessing. Findings without evidence are out of scope, and every reported issue must cite a file path, line range, and constitution rule or requirement identifier.