revision-external-api
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through data retrieved from the Revision External API (e.g., component names, descriptions, diagram metadata).
- Ingestion points: Data retrieved via GET requests to
/api/external/components,/api/external/diagrams, and/api/external/typesas defined in SKILL.md and OPENAPI.md. - Boundary markers: The skill does not define specific delimiters or instructions to prevent the agent from following commands that might be embedded in the retrieved architecture data.
- Capability inventory: The skill utilizes
curlfor network operations and generates local YAML templates, creating a pathway where malicious API data could influence subsequent API writes or local file content. - Sanitization: There is no mention of sanitizing or escaping content retrieved from the API before it is processed by the agent or included in generated templates.
Audit Metadata