Use `review-agent` for an ordinary Engineering Brief, Task Plan, acceptance, or contract artifact. Route high-risk design and release, deployment, or migration readiness to their specialist reviewers.
Scanned 9/8/2026
Install to Claude Code
npx -y skills add machenjie/rd-skills --skill engineering-artifact-review --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Engineering Artifact Review?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/machenjie-engineering-artifact-review)More formats (shields.io, HTML) on the badges page.
---
name: engineering-artifact-review
description: "Use `review-agent` for an ordinary Engineering Brief, Task Plan, acceptance, or contract artifact. Route high-risk design and release, deployment, or migration readiness to their specialist reviewers."
---
# Engineering Artifact Review
## Role
Support `review-agent` in assessing a concrete engineering artifact's
decisions, dependencies, failure boundaries, and evidence.
## When To Use
- ordinary Engineering Brief Task Plan acceptance or contract artifact needs independent review
- pre-implementation decision affects downstream tasks
## Do Not Use
- actual implementation diff review; use `ai-code-review-refactor` or a risk-specific Skill
- an ordinary Direct Task has no separate decision artifact
- high-risk Engineering Brief; use `high-risk-design-review`
- release deployment or migration readiness; use `delivery-release-gate`
## Required Inputs
- observable acceptance and non-goals
- actual artifact and source evidence it cites
- decision, ownership, dependency, validation, and rollback boundaries
## Professional Decision Rules
- Verify every material claim against current source or explicit owner evidence.
- Check owner, invariants, reuse, public contracts, data, failure
behavior, dependencies, validation, and rollback proportional to risk.
- Separate reviewed, unreviewed, and unverified scope.
- Report findings with severity, affected decision, downstream impact, and required action.
## High-Value Gotchas
- A complete template can still encode the wrong owner.
- A Task Plan without a safe first slice can prolong preparation indefinitely.
- Rollback prose is not an executable rollback path.
## Execution Checklist
1. Confirm the artifact, acceptance, owner, and review boundary.
2. Verify cited source and rejected alternatives.
3. Check downstream dependencies, collision risks, integration, and validation.
4. Check failure behavior, compatibility, rollback, and user-owned decisions.
## Stop / Escalation Conditions
- Stop when the artifact or its material source evidence is unavailable.
- Route a high-risk Engineering Brief to `high-risk-design-review` without issuing its specialist verdict here.
- Route release, deployment, or migration readiness to `delivery-release-gate` without issuing a go/no-go verdict here.
- Escalate destructive, production, privileged, irreversible, security, privacy,
money, or public-contract choices without authoritative decisions.
## Output Contract
- verdict and severity-ranked findings
- reviewed, unreviewed, and unverified scope
- owner, dependency, first-slice, validation, and rollback assessment
- downstream impact and residual risk
## Targeted References
| Path | Type | Load when | Do not load when | Required by | Required output |
|---|---|---|---|---|---|
| [review](references/review-checklist.md) | decision-checklist | ordinary pre-implementation artifact decisions have downstream impact | the task is implementation diff review, a Direct Task, high-risk design, or release/deployment/migration readiness | review-agent | checklist-result, residual-risk |
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!