suggest

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes user-provided code and intent to suggest relevant Convex components, which constitutes a standard ingestion of untrusted data.
  • Ingestion points: The skill monitors codeSnippets and userAsk from the user context.
  • Boundary markers: No specific delimiters or boundary instructions are defined in the skill for isolating user code.
  • Capability inventory: The skill is authorized to surface suggestions and invoke the /add command for component installation only after receiving explicit user confirmation.
  • Sanitization: No explicit sanitization or filtering of the ingested code snippets is described; the skill relies on pattern matching for component detection.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 02:52 AM
Security Audit — agent-trust-hub — suggest