github-ai-features-2025
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for GitHub Actions that trigger on external events like issue labels. These workflows ingest metadata from issues which could be controlled by an external actor to influence the AI agent's behavior during analysis or code generation.
- Ingestion points: Issue number and title are used in the ai-bugfix.yml workflow defined in SKILL.md.
- Boundary markers: The provided workflow examples do not include explicit sanitization or boundary markers for the issue content.
- Capability inventory: The documented workflows use the github/ai-agent tool which is capable of generating code and opening Pull Requests.
- Sanitization: No input sanitization is present in the example YAML configurations.
Audit Metadata