gitlab-code-reviewer

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly aligned with GitLab code review, but it can post MR comments without a clear confirmation step and may send your GitLab token to the host contained in a supplied MR URL.

Use this only with a dedicated, least-privileged GitLab token. Before running it, verify that MR URLs are on the exact GitLab host you trust, and do not allow it to post comments unless you have reviewed and approved the exact comments.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

#
ASI03: Identity and Privilege Abuse
High
What this means

Your GitLab personal access token could be exposed to a server named in an MR URL if that URL is not the intended GitLab host.

Why it was flagged

The GitLab token is loaded from the local credential file, but API requests use the host parsed from the user-supplied MR URL rather than validating it against the configured credential host. A crafted or mistaken MR URL could cause the token to be sent to an unintended host.

Skill content
creds = load_credentials()
host, project, iid = parse_mr_url(mr_url)
...
data = api_get(host, creds["token"], ...)
...
result = api_post(host, creds["token"], ...)
Recommendation

Validate that the MR URL host exactly matches the configured credential host, or require separate per-host credentials and refuse mismatches before making any API request.

#
ASI02: Tool Misuse and Exploitation
Medium
What this means

The agent could post review comments to a real merge request based on its own generated findings, which may affect teammates or project workflow.

Why it was flagged

Posting comments is described as a normal workflow step once write scope is available. The artifact does not clearly require a separate user confirmation or an explicit 'post these comments' request before mutating the MR.

Skill content
### 6. Post inline comments to GitLab

Only execute this step if `check-token` (step 1) returned `"can_write": true`.
...
python scripts/post_comments.py <mr_url> /tmp/mr_comments.json
Recommendation

Default to chat-only review, show the exact comments to be posted, and require explicit user approval before running the posting script.

#
ASI03: Identity and Privilege Abuse
Low
What this means

Installing and using the skill may require storing a GitLab token with read or write access to projects the token can reach.

Why it was flagged

The skill explicitly uses a GitLab personal access token, including the broad 'api' scope for posting. This is purpose-aligned for GitLab review, but users should notice the required account authority.

Skill content
Credentials: `~/.openclaw/credentials/gitlab.json`
...
Required API scopes:
- `api` — required for posting inline comments
- `read_api` — sufficient for analysis only
Recommendation

Use the least-privileged token possible, prefer read_api for analysis-only reviews, and create a dedicated token limited to the intended GitLab host and projects.