Agent skill · Audited from source
pr-review
skill · minimax-ai/skillsMIT licenceAudited 31 Oct 2026 · source ceddde63e832
Is it safe? With care — CAUTION, grade B (89/100). We read this skill's source at commit ceddde63e832 and found no filesystem, network or shell access and no hard-coded secrets. The audit raised 1 finding, 1 critical or high. What it does: Review pull requests for the MiniMax Skills repository. Use when reviewing PRs, validating new skill submissions, or checking existing skills for compliance. Run the validation script first for hard checks, then apply…
Cite
Cite this page
ModelsAtlas. “pr-review (minimax-ai/skills)”. Retrieved 31 Oct 2026. https://modelsatlas.com/skills/minimax-ai/pr-review
6 skills are called “pr-review”. You are viewing minimax-ai/skills.
- minimax-ai/skills★ 14k!CAUTION
- pytorch/pytorch★ 105k!CAUTION
- rtk-ai/rtk★ 82k✓SAFE
- paperclipai/companies★ 919!CAUTION
- aaronjmars/aeon★ 774✓SAFE
and 1 more with this name.
Trust audit
Audit 0.4.1 · source read 31 Oct 2026
CAUTION means the audit found things worth knowing before you trust this skill. Grade B is shared by 16,667 of 17,602 audited skills.
What the source does
- FilesystemNone observedNo reads or writes in the source
- NetworkNone observedNo outbound calls in the source
- ShellNone observedNo commands run in the source
- DependenciesPinnedNothing floats
- SecretsNone foundNo keys in source
Five audit layers
| Layer | What we check | Result | Findings |
|---|---|---|---|
L0 Provenance | Repo owner, licence, archive status, activity | PASS | 0 |
L1 Static source | Secrets, path traversal, supply chain, hidden code | PASS | 0 |
L2 Instructions | Prompt injection, authority framing, hidden directives | FAIL | 1 |
L3 Inventory | The files and declarations match the source | PASS | 0 |
L4 Sandbox | Run in isolation and watch behaviour | SKIPPED why | 0 |
Gates applied: no behavioural pass.
Findings (1)
HIGHPrompt injectionprompt.credential_read · SKILL.md:44
4. **API key handling** — If external APIs are used, are credentials read from environment variables?
- Why
- asks the agent to read credentials
- Fix
- —
Audit 0.4.1 · source sha ceddde63e832 · 31 Oct 2026Report an issue or request a re-scan
Provenance
Who published it, and what the verdict is pinned to
- Publisherminimax-ai GitHub organizationOwner of the repository this skill is read from
- Source folderminimax-ai/skills/.claude/skills/pr-reviewThe exact folder we audited
- LicenceMITAs the repository states it
- Commit pinned for this verdictceddde63e832A new audit runs when the source changes
- Repository statusNot archivedLast push 195 days ago
- Sandboxed runNot runSkills are read, not executed; layer L4 is skipped
What it tells the agent
Full source on GitHub ↗Its description: Review pull requests for the MiniMax Skills repository. Use when reviewing PRs, validating new skill submissions, or checking existing skills for compliance. Run the validation script first for hard checks, then apply…
--- name: pr-review description: > Review pull requests for the MiniMax Skills repository. Use when reviewing PRs, validating new skill submissions, or checking existing skills for compliance. Run the validation script first for hard checks, then apply quality guidelines for content review. Triggers: PR review, pull request, validate skill, check skill. license: MIT metadata: version: "1.0" category: tooling --- # PR Review Skill Review pull requests against repository standards. Two-phase process: automated validation, then manual content review. ## Phase 1: Automated Validation (Hard Rules) Run the validation script to check structural requirements: ```bash python .claude/skills/pr-review/scripts/validate_skills.py ``` The script checks: - `SKILL.md` exists in every skill directory - YAML frontmatter is parseable - Required fields present: `name`, `description` - `name` matches directory name - No hardcoded secrets detected All ERROR-level checks must pass. WARNING-level items (missing `license`, `metadata`) should be flagged but are not blockers. See [references/structure-rules.md](references/structure-rules.md) for the complete hard rules specification. ## Phase 2: Content Review (Soft Guidelines) After automated checks pass, review the PR against quality guidelines: 1. **Skill scope** — Does it overlap with existing skills? Is the boundary clear? 2. **Description quality** — Does the `description` include clear trigger conditions? 3. **File size** — Are reference docs reasonably sized for context window consumption? 4. **API key handling** — If external APIs are used, are credentials read from environment variables? 5. **Script quality** — Do scripts have shebang, requirements.txt, and error handling? 6. **Language** — Are SKILL.md and code written in English? 7. **README sync** — Are `README.md` and `README_zh.md` updated for new skills? See [references/quality-guidelines.md](references/quality-guidelines.md) for soft guidelines details. ## Review Checklist Summary ### Must Pass (Blockers) - [ ] `validate_skills.py` exits with code 0 - [ ] PR title follows conventional commit format - [ ] One PR, one purpose ### Should Pass (Flagged in Review) - [ ] No functional overlap with existing skills - [ ] Description includes trigger conditions - [ ] Files are reasonably sized - [ ] API keys via environment variables - [ ] README tables updated for new skills (Source column set to `Community`)
Excerpt, first 2,461 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.
Install
Pinned to the audited commit ceddde6
git clone https://github.com/minimax-ai/skills git -C skills checkout ceddde63e832
The author documents no install command, so this uses the repository directly. Our verdict covers commit ceddde63e832 only.
Works in
The author has not said which agents this skill supports, and we have not tested it in any.
Audit history
Every audit, every source change
- !CAUTION B 89 / 100
first audit · source
ceddde63e832
README badge
For maintainers · shows the latest audit
[](https://modelsatlas.com/skills/minimax-ai/pr-review/)
The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.
Similar skills
All coding & review skills →obra/superpowers
Use when creating new skills, editing existing skills, or verifying skills work before deployment
mattpocock/skills
A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.
mattpocock/skills
Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
Add this badge to your README
[](https://modelsatlas.com/skills/minimax-ai/pr-review/)
<a href="https://modelsatlas.com/skills/minimax-ai/pr-review/"><img src="https://modelsatlas.com/skills/minimax-ai/pr-review/badge.svg" alt="ModelsAtlas: CAUTION B"></a>
Questions about pr-review
Is pr-review safe to install?
What can pr-review access on my machine?
How current is this page?
Sources
- github.com/minimax-ai/skills/tree/main/.claude/skills/pr-review — the audited source at commit ceddde63e832 and every finding aboveObserved · 31 Oct 2026
Spotted something wrong? Report a correction · Methodology
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- Is pr-review safe to install?
- What can pr-review access on my machine?
- How current is this page?