frontend-logic-integration
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data sources (BFF/Sitecore manifests) and uses the file-write tool to generate source code, creating a potential vector for indirect prompt injection. 1. Ingestion points: The skill consumes BFF contracts, Sitecore contracts, and prop models provided in the manifest to drive the generation steps in SKILL.md. 2. Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded instructions when interpolating manifest data into generated TypeScript/React files. 3. Capability inventory: The skill employs the file-write tool extensively to create files across multiple project directories (Components, Hooks, Services, etc.). 4. Sanitization: Absent; the instructions do not mandate validation, escaping, or filtering of the manifest content before it is written to the filesystem.
- [COMMAND_EXECUTION]: The skill requires the agent to utilize a file-write tool to generate and persist multiple TypeScript and React source files to the local disk based on provided specifications.
Audit Metadata