abridge-rate-limits
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill utilizes the WebFetch tool to ingest content from external documentation sites including abridge.com and hhs.gov. This creates a potential surface for indirect prompt injection where malicious instructions could be embedded in the documentation retrieved at runtime.
- Ingestion points: The references/official-docs.md file defines several external URLs that the agent is instructed to fetch and analyze during the capacity planning workflow.
- Boundary markers: The skill lacks technical delimiters or explicit instructions to ignore embedded commands within the fetched content, although it does instruct the agent to separate verified facts from assumptions.
- Capability inventory: The skill possesses the Write and Edit tools, which allow the agent to modify local repository configuration and test suites based on information derived from the external documentation.
- Sanitization: The instructions do not specify any automated sanitization, integrity validation, or filtering of the fetched HTML/Markdown content before it is processed by the agent.
Audit Metadata