code-reviewer

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a code-review guide, but it also asks for broad team/repository metrics using local repo paths and token-protected PR APIs that go beyond normal Go code review.

Install only if you want both code review and the included team-metrics behavior. Before use, restrict it to the repo you choose, avoid giving broad access tokens, and approve any test execution or PR-metrics collection explicitly.

VirusTotal

41/41 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
Medium
What this means

The agent may request or use a project access token and retrieve organization PR/review metrics that are not needed for ordinary code review.

Why it was flagged

The skill tells the agent to use token-protected PR review statistics APIs, while the registry declares no credentials and the main purpose is code review.

Skill content
认证方式: Bearer Token ... 需要有效的访问令牌(Access Token)
Recommendation

Only use this part if you intentionally want team metrics; require a least-privilege read-only token, explicit user approval, and a clearly scoped project and time range.

#
ASI02: Tool Misuse and Exploitation
Medium
What this means

If those paths exist, the agent could inspect unrelated local corporate repositories and team activity as part of a code-review workflow.

Why it was flagged

The rule directs evaluation across multiple hard-coded local repositories rather than a user-selected project, broadening local data access.

Skill content
评估团队代码仓库本地地址

 C:\yanfayun\gpc-srv
 C:\yanfayun\gpc-provider-orchestrator
 ...
 C:\yanfayun\provider-srv
Recommendation

Remove hard-coded repository paths and require the user to explicitly choose the repository, branch, date range, and metrics to analyze.

#
ASI05: Unexpected Code Execution
Low
What this means

On an untrusted PR or repository, test execution could run code you did not intend to execute.

Why it was flagged

Running Go tests is purpose-aligned for coverage analysis, but it executes code from the reviewed repository and writes coverage artifacts.

Skill content
Agent MUST execute this command to generate coverage data:

go test "./..." ... -json > test-report.json
Recommendation

Run coverage commands only with user approval, preferably in a sandbox or trusted development environment.