semantic-search

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): The skill contains no executable scripts or command-line instructions. It primarily consists of high-level principles and regex-based validation rules for a linter.
  • [CREDENTIALS_UNSAFE] (SAFE): The file references/validations.md contains a rule to detect exposed API keys (exposed-api-key). This is a security-positive detection pattern, not a credential leak itself.
  • [DATA_EXFILTRATION] (SAFE): No network operations or sensitive file access patterns were found. The mentioned services (Pinecone, Cohere, etc.) are standard industry tools referenced for architectural context.
  • [PROMPT_INJECTION] (SAFE): The instructions do not contain any bypass attempts or overrides of agent behavior. The 'Reference System Usage' section correctly defines grounded sources of truth for the agent.
  • [INDIRECT_PROMPT_INJECTION] (LOW): While the skill describes systems that process untrusted user queries, it explicitly includes a validation rule (no-input-validation) to encourage developers to sanitize and validate search queries.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:14 PM