github-issue-writing
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user input to generate GitHub issue titles and bodies. * Ingestion points: The agent takes user-supplied descriptions and bug details to populate issue templates in SKILL.md. * Boundary markers: There are no explicit delimiters or instructions to the model to ignore potential commands embedded within the user-provided issue content. * Capability inventory: The skill has access to powerful tools including Bash(gh issue *), Bash(gh label *), and gh api calls via the allowed-tools configuration in the frontmatter. * Sanitization: While the skill correctly recommends using --body-file to prevent shell-level command injection and formatting errors, it does not sanitize the logical content of the issue body to prevent the agent from being manipulated by instructions hidden in the text (e.g., in HTML comments or markdown blocks).
Audit Metadata