update
Fail
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (HIGH): The skill's primary function is to download content from an untrusted GitHub user ('boshu2'). This user is not on the list of trusted entities, making the source unverified.
- [REMOTE_CODE_EXECUTION] (HIGH): The command
npx skills@latest add boshu2/agentops --all -gdownloads a package from the npm registry and immediately uses it to install and potentially execute scripts from a remote repository. This constitutes remote code execution (RCE) via an untrusted third party. - [COMMAND_EXECUTION] (MEDIUM): The skill executes commands with global impact (
-gflag) and suggests manual file copying (cp -r) into the sensitive~/.claude/skills/directory. This bypasses standard safety checks and allows for persistent modification of the agent's environment.
Recommendations
- AI detected serious security threats
Audit Metadata