career
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands including
git,find,gh(GitHub CLI), andjqwithinreferences/staff-engineer-progress.mdto gather engineering metrics. These commands are used to quantify technical writing, code review activity, and module ownership for the purpose of generating a career progress report. The commands are read-only and restricted to local repository metadata and GitHub activity. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it processes untrusted data from the local repository and GitHub.
- Ingestion points: Untrusted data enters the agent context through commit messages, file names, and pull request descriptions retrieved from the repository history in
references/staff-engineer-progress.md. - Boundary markers: The instructions do not define specific delimiters or warnings for the agent to ignore instructions potentially embedded in this metadata.
- Capability inventory: The skill's capabilities are limited to reading repository data and generating text reports; it lacks file-writing or external network exfiltration capabilities (other than standard GitHub API calls via
gh). - Sanitization: There is no explicit sanitization or filtering of the repository metadata before it is processed into the final report.
Audit Metadata