enhance-web-instant-nav
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could potentially contain malicious instructions designed to influence agent behavior.
- Ingestion points: The skill reads output from shell commands such as npx @lhci/cli (Lighthouse CI) and curl (HTTP headers) in Phase 1 of SKILL.md.
- Boundary markers: Absent. No delimiters are defined to separate tool output from the skill's own instructions.
- Capability inventory: The agent can execute network and shell commands (curl, npx) and is expected to write configuration blocks and code diffs.
- Sanitization: Absent. There is no evidence of validation or sanitization of the external tool output before processing.
- [EXTERNAL_DOWNLOADS]: The skill uses npx to execute external packages from a public registry.
- Evidence: The command npx @lhci/cli autorun is specified in SKILL.md. Note: The package is maintained by a well-known service.
Audit Metadata