howto-code-in-rust
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill mentions using external research tools to gather information about crates and repositories.
- Ingestion points: Instructions suggest using vendor-owned tools (ed3d-research-agents:internet-researcher and ed3d-research-agents:remote-code-researcher) to fetch data from the internet and external repositories.
- Boundary markers: The skill specifies that these research tools run in isolated contexts and return summaries to the main agent.
- Capability inventory: The skill is focused on writing, reviewing, and modifying Rust source code.
- Sanitization: Content is summarized before being incorporated into the agent's working context, which serves as a filtering layer for external data.
Audit Metadata