ash-framework
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data by instructing the agent to fetch documentation from online and local sources.
- Ingestion points: The skill utilizes
WebFetchto access hexdocs.pm andmcp__tidewave__get_docsfor module-specific reference material (SKILL.md). - Boundary markers: Instructions include specific prompt parameters (e.g., "Extract module docs.") to delimit the extraction of information from the retrieved documentation.
- Capability inventory: The agent is guided to generate Elixir code, modify resource definitions, and execute framework-specific
mixcommands. - Sanitization: Content retrieved from documentation sources is processed by the agent without additional sanitization layers described in the skill, relying on standard model safety boundaries.
Audit Metadata