write-an-issue
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is the creation of local documentation. It uses a structured interview process to gather requirements and format them into a markdown template.
- [COMMAND_EXECUTION]: The instructions explicitly forbid the use of external commands or the GitHub CLI, ensuring the agent only performs file system writes to a specific local directory (
issues/). - [EXTERNAL_DOWNLOADS]: The skill does not reference or download any external scripts, packages, or remote resources.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to define issue behaviors. While it processes untrusted data (user descriptions) and writes it to files without explicit sanitization or boundary markers, the impact is limited to local text documentation with no follow-on execution of that data.
Audit Metadata