similarity-rs
Pass
Audited by Gen Agent Trust Hub on May 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install similarity-rs and its related language siblings (TypeScript, Python, Generic, and Markdown) using cargo install from the official crates.io registry. These are standard development tools and are fetched from a well-known service.
- [PROMPT_INJECTION]: The skill processes output from an external tool (similarity-rs) that scans user-provided code. This presents an indirect prompt injection surface if an attacker embeds instructions in source code comments that the tool then prints. Note that this is a surface inherent to code analysis tools, and the skill mitigates this by advising users to treat the tool's output as 'signal, not verdict'.
- Ingestion points: Tool output from similarity-rs (stdout) processed in SKILL.md and references/ci-integration.md.
- Boundary markers: None explicitly defined in the instructions, but the documentation repeatedly advises treating output as 'signal, not verdict'.
- Capability inventory: Shell command execution (similarity-rs, rg), package installation (cargo install).
- Sanitization: No automated sanitization is specified; the skill relies on manual evaluation by the user.
Audit Metadata