How do I install the Zr Validate skill?

Run npx skillmds@latest add zenon-red/zr-validate in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Zr Validate skill do?

Validate one peer review outcome when dispatched a ValidateReview action. It is listed under Coding & Dev Tools on SkillMD.

Is Zr Validate safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Zr Validate?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Zr Validate free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Zr Validate?

zenon-red (@zenon-red) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Zr Validate by zenon-red — Agent Skill

Name: Zr Validate
Author: zenon-red

File contents

zr-validate

Job

Perform independent validation for one completed peer review.

Inputs

dispatched action payload with kind: ValidateReview
routed review/task/PR context

Validation Requirements

For the routed review, verify all three:

Reviewer independence
- Reviewer is not the task owner/PR author
- Validation is by a different agent than the reviewer
Outcome is grounded in evidence
- Review all review comments
- Check for "Critical" severity issues
- Ensure no "Request changes" with critical/blocking issues
Summary is substantive
- Not just "LGTM" rubber-stamps
- Reviews should reference task requirements
- At least some code quality feedback (even if minor)

Validation Checklist

PR: [URL]
Task: #[ID]
Review: #[REVIEW_ID]

Validation:
- [ ] Reviewer independence verified
- [ ] Outcome matches evidence
- [ ] Summary is substantive

If all checked: record validation as valid
If any failed: record validation as invalid with rationale

Steps

Load routed review/task/PR context.
Inspect review summary and supporting PR context.
Apply the three validation checks above.
Decide valid or invalid.
Leave a GitHub review/comment artifact, then complete via Probe:

probe action show <action-id> --json
probe task get <task-id>
gh pr view <number> --repo <org.github_org>/<repo> --json reviews,comments
probe review validate <action-id> --outcome valid --summary "<why>" --artifact-url <github-comment-or-review-url>
# or --outcome invalid when checks fail

What This Skill Does NOT Do

Deep code review
Task ownership assignment
Final merge execution

This skill validates a routed review and records that validation as action completion.

Anti-Patterns

Validating your own review
Ignoring obvious blocking concerns
One-line rubber-stamp validation summaries
Completing validation without checking routed context

Output Contract

Routed ValidateReview action is completed with valid or invalid.
Validation summary is explicit and audit-friendly.

Merge-ready dispatch (`MergeReadyTask`)

When routed a merge-ready task action instead of validation:

probe action show <action-id> --json
probe task get <task-id>
# merge the linked PR or otherwise finalize the task, then:
probe action complete-merge <action-id>

Do not call probe action complete for merge-ready routes.

---name: zr-validatedescription: Validate one peer review outcome when dispatched a ValidateReview action.---# zr-validate## JobPerform independent validation for one completed peer review.## Inputs- dispatched action payload with `kind: ValidateReview`- routed review/task/PR context## Validation Requirements**For the routed review, verify all three:**1. **Reviewer independence** - Reviewer is not the task owner/PR author - Validation is by a different agent than the reviewer2. **Outcome is grounded in evidence** - Review all review comments - Check for "Critical" severity issues - Ensure no "Request changes" with critical/blocking issues3. **Summary is substantive** - Not just "LGTM" rubber-stamps - Reviews should reference task requirements - At least some code quality feedback (even if minor)## Validation Checklist```PR: [URL]Task: #[ID]Review: #[REVIEW_ID]Validation:- [ ] Reviewer independence verified- [ ] Outcome matches evidence- [ ] Summary is substantiveIf all checked: record validation as validIf any failed: record validation as invalid with rationale```## Steps1. Load routed review/task/PR context.2. Inspect review summary and supporting PR context.3. Apply the three validation checks above.4. Decide `valid` or `invalid`.5. Leave a GitHub review/comment artifact, then complete via Probe:```bashprobe action show <action-id> --jsonprobe task get <task-id>gh pr view <number> --repo <org.github_org>/<repo> --json reviews,commentsprobe review validate <action-id> --outcome valid --summary "<why>" --artifact-url <github-comment-or-review-url># or --outcome invalid when checks fail```## What This Skill Does NOT Do- Deep code review- Task ownership assignment- Final merge execution**This skill validates a routed review** and records that validation as action completion.## Anti-Patterns- Validating your own review- Ignoring obvious blocking concerns- One-line rubber-stamp validation summaries- Completing validation without checking routed context## Output Contract- Routed ValidateReview action is completed with `valid` or `invalid`.- Validation summary is explicit and audit-friendly.## Merge-ready dispatch (`MergeReadyTask`)When routed a merge-ready task action instead of validation:```bashprobe action show <action-id> --jsonprobe task get <task-id># merge the linked PR or otherwise finalize the task, then:probe action complete-merge <action-id>```Do not call `probe action complete` for merge-ready routes.

Frequently asked questions

Zr Validate