fork-intelligence
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources (GitHub forks) which may contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: SKILL.md (Steps 2, 5, 6, 7) and references/signal-priority.md contain instructions to fetch fork metadata, commit messages, branch names, tags, and pull request titles using the
gh apitool. - Boundary markers: The instructions do not define clear delimiters or use "ignore embedded instructions" warnings when presenting this external data to the agent for evaluation.
- Capability inventory: The skill utilizes Bash (to execute
gh api), Grep, Glob, and Read tools. While the Write tool is not explicitly in the allowed-tools list, the instructions command the agent to "fix this file immediately" if errors are found, indicating a design that encourages self-modification of instructions based on runtime experience. - Sanitization: No sanitization, escaping, or validation of the fetched GitHub content is performed before it is added to the agent's context.
- [EXTERNAL_DOWNLOADS]: The skill fetches metadata and code history from GitHub.
- Source: Interacts with the official GitHub API via the
ghCLI tool. - Content: Accesses public repository information including fork details, commit history, and pull request data as part of its core analytical function.
Audit Metadata