create-ol-github-issue
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input to populate GitHub issue fields, creating a potential surface for indirect prompt injection or command manipulation.
- Ingestion points: User-provided strings for the repository name, issue title, and various body sections (Expected Behavior, Steps to Reproduce, etc.) as described in SKILL.md.
- Boundary markers: The instructions include a mandatory human-in-the-loop confirmation step in Step 3 ("Show the filled-in body and confirm before creating it").
- Capability inventory: Execution of the
gh issue createcommand via subprocess in SKILL.md. - Sanitization: No explicit sanitization or escaping of shell metacharacters is defined for the user-supplied input.
- [COMMAND_EXECUTION]: The skill constructs and executes a shell command using the GitHub CLI (
gh). It dynamically inserts user input into the--repo,--title,--body, and--labelflags. While the use of quotes in the template reduces risk, the lack of explicit character escaping for shell metacharacters in the user input constitutes a minor security concern.
Audit Metadata