write-issue
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill interacts with the GitHub CLI (
gh) and a bundled Python script to manage issues. - The implementation uses safe practices such as writing content to temporary files using
mktempand heredocs with quoted delimiters (<<'EOF') to prevent shell expansion. - The Python script
scripts/audit_issue_bodies.pyusessubprocess.check_outputwith a list of arguments (shell=False), which is a secure method for executing external commands and prevents command injection. - [DATA_EXFILTRATION]: Network operations are restricted to the official GitHub CLI and standard repository interaction. There is no evidence of data being sent to unauthorized third-party domains or sensitive local files being accessed.
- [PROMPT_INJECTION]: The skill provides robust instructions to sanitize content when rewriting issues. It explicitly directs the agent to remove local workspace paths, session labels, and unexplained shorthand, which mitigates risks associated with processing untrusted data from existing issue bodies.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute any external scripts or packages. It relies solely on pre-installed system tools (
gh,python3) and its own included audit script.
Audit Metadata