Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Gitlab Mr Review
v1.0.0Automatically fetches and reviews GitLab merge requests using a template to identify issues and post detailed, actionable review comments.
⭐ 0· 154·2 current·2 all-time
byzachary.wu@wujinyuan
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill claims to review GitLab MRs but the manifest declares no required binaries or credentials while the runtime instructions explicitly require the 'glab' CLI and rely on glab-stored GitLab credentials. This omission is inconsistent: a MR-reviewer skill normally must declare/globally require the glab binary (or equivalent API credentials).
Instruction Scope
SKILL.md outlines a narrow, sensible workflow (parse MR URL, fetch MR and changes, analyze diff using the provided template, post a comment). It does not instruct reading unrelated system files or exfiltrating data. However, it references posting content from 'review-comment.md' (cat review-comment.md) though that file is not included; it also assumes glab is configured with credentials (not documented).
Install Mechanism
Instruction-only skill with no install spec or remote downloads — lowers risk. Nothing is written to disk by an installer. The only runtime risk comes from executing glab commands, which are invoked at runtime (not installed by the skill).
Credentials
The skill declares no environment variables or primary credential, yet it relies on glab which uses local configuration (including stored GitLab tokens). This is proportionate in purpose but the lack of explicit declaration/documentation about required credentials or how auth is handled is an omission that could surprise users.
Persistence & Privilege
always is false and there is no install-time persistence. The skill can be invoked by agents autonomously (platform default), which is expected for a review tool, but users should be aware it will post comments to MRs when run.
What to consider before installing
This skill appears to do what it says (fetch MR diffs and post review comments), but there are important omissions and small inconsistencies you should address before using it:
- glab dependency: The SKILL.md runs 'glab' commands but the manifest does not declare that the glab CLI is required. Ensure glab is installed and tested on the agent host before using this skill.
- Authentication: glab uses local configuration (personal access tokens or saved credentials). The skill does not declare or request credentials; verify which GitLab account the agent will use and whether you are comfortable letting it post comments using that identity.
- Missing output file: The instructions post body from 'review-comment.md' (cat review-comment.md), but the repository only includes 'code-review-template.md' and not 'review-comment.md'. Confirm how the review content is generated/saved before posting (the template exists, but the exact write/read workflow is unclear).
- Safety: The skill will post comments to real MRs when invoked. If you want to test safely, run it against a test repository or ensure the agent runs with read-only credentials until you trust the behavior.
Recommendation: Ask the publisher (or edit the skill) to (1) declare 'glab' as a required binary, (2) document how authentication is handled (what credentials are needed and where they come from), and (3) fix the mismatch between 'review-comment.md' and the included template file. With those clarifications the skill would be coherent and likely benign; as-is it is sloppy and requires caution.Like a lobster shell, security has layers — review code before you run it.
latestvk976b68as6g0gz1a8d7sqfe9nd833sny
License
MIT-0
Free to use, modify, and redistribute. No attribution required.