client-data-fetching
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read local project lockfiles to verify the version of
@tanstack/react-query. This creates a data ingestion surface where the agent processes content from local repository files. Additionally, the skill provides guidance on handling external server error data, specifically recommending sanitization to prevent data leakage. - Ingestion points: Reads version information from repository lockfiles (e.g., package-lock.json, pnpm-lock.yaml).
- Boundary markers: None explicitly defined for the file reading process.
- Capability inventory: Uses the
repo-filestool to access and read repository content. - Sanitization: The skill explicitly instructs the user/agent to sanitize server error bodies before surfacing them to UI components.
- [SAFE]: The skill references external resources on GitHub and npm that are associated with the author's account (
bh611627). These links represent documentation and package references consistent with the skill's stated purpose.
Audit Metadata