audit-agent-context
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local git commands (
git rev-parse,git log,git worktree list) viaexecFileSyncinscripts/checks.mjsto analyze project history and structure. This is a legitimate diagnostic function of the tool and is limited to local repository metadata. - [REMOTE_CODE_EXECUTION]: The skill provides instructions to run a bundled Node.js script (
node <this-skill-dir>/scripts/audit-bloat.mjs). Analysis of the script confirms it only performs local file system reads and standard output reporting for the purpose of auditing context bloat. - [SAFE]: The skill does not perform any network operations, does not access sensitive credential files, and does not attempt to persist itself or escalate privileges. Its operations are restricted to measuring and reporting on the markdown and configuration files within the provided project directory.
Audit Metadata