skills.rest
j5ik2o
j5ik2oCOMMUNITY
Explore all skills by this author →
Website𝕏@j5ik2oGitHub

kiro-review

Reviews task implementations against approved specs, boundaries, and verification evidence.

Updated Jun 23, 2026
Search Tags:#code review#quality gate#git diff#test evidence#task verification#spec validation#boundary audit
One-click install
npx skills add https://github.com/j5ik2o/marp-ai-base --skill kiro-review-j5ik2o
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: kiro-review
Source: https://github.com/j5ik2o/marp-ai-base/tree/main/.agents/skills/kiro-review
Command: npx skills add https://github.com/j5ik2o/marp-ai-base --skill kiro-review-j5ik2o

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents incomplete, out-of-scope, or unverified code changes from being accepted by performing adversarial task-level review against approved requirements, design, and mechanical evidence. ## Core Features & Use Cases - Mechanical Verification: Runs the project's test suite, greps for placeholder markers and hardcoded secrets, and checks boundary compliance before judging code. - Spec and Design Alignment: Confirms the implementation satisfies referenced requirement and design sections without silent substitutions or scope drift. - Structured Verdicts: Returns APPROVED or REJECTED with findings, severity levels, and mandatory remediation steps. - Use Case: After an implementer marks a task READY_FOR_REVIEW, run this review to inspect the git diff, verify RED phase evidence, and reject changes that spill outside the task's declared boundary. ## Quick Start Review task 3.2 from the current spec against its requirements, design sections, and boundary constraints, then give me the verdict.

Frequently Asked Questions about kiro-review

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

FAQPage Schema
How do I review a task implementation against a spec?▼

Provide the task ID, relevant requirement and design section numbers, spec file paths, and the implementer's status report. The review runs git diff first, executes mechanical checks like tests and boundary comparison, then returns an APPROVED or REJECTED verdict with findings.

What checks does an adversarial code review run before approval?▼

It runs the project's test suite, greps changed files for TBD/TODO/FIXME markers and hardcoded secrets, compares changes against the task boundary scope, verifies RED phase evidence for behavioral tasks, and runs lint or targeted static spot checks.

When should a task review be rejected instead of approved?▼

Reject when tests fail, placeholder markers or secrets are introduced, changes spill outside the declared boundary, RED evidence is missing, or the implementation silently deviates from approved requirements or design-mandated structures.

Can a review pass just because all tests pass?▼

No. Passing tests do not prove spec compliance or boundary respect. The review also checks acceptance criteria coverage, requirements and design alignment, test quality, error handling, and hidden cross-boundary coupling.

What happens when the spec is ambiguous during review?▼

The review escalates instead of guessing when the spec is ambiguous in a correctness-critical way, the design conflicts with technical reality, required evidence cannot be gathered, or boundary ownership cannot be determined cleanly.

Related Skills

Explore More Software Engineering Skills→
notta50notta50

kiro-spec-tasks

Generate implementation tasks from feature specs into structured task graphs.

Community
Advanced
pinkroosteraipinkroosterai
1

pm-done

Mark tasks, issues, and feature requests as completed with cascading state changes.

Community
Advanced
ZaxbyHubZaxbyHub

qa-gate-disciplined-completion

Verify QA gate artifacts, tests, and lint before marking tasks complete.

Official
Intermediate
TumeloRamaphosaTumeloRamaphosa

review

Analyzes pull request diffs for security and structural risks before merge.

Community
Advanced