pbvex-client

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides explicit security warnings to prevent the inclusion of superuser or deployment tokens in client-side code, which is a standard best practice for client-server architectures.
  • [COMMAND_EXECUTION]: The skill utilizes local development tools (sed, rg, pnpm) to inspect the project's source code and run internal tests. These operations are restricted to the local development environment and serve the skill's primary purpose of implementation and troubleshooting.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes the ingestion of data from a backend via client.query.
  • Ingestion points: Data returned from PBVex backend queries described in SKILL.md.
  • Boundary markers: None explicitly defined in the prompt templates.
  • Capability inventory: Local file reading (sed, rg) and test execution (pnpm) in SKILL.md.
  • Sanitization: Not explicitly mentioned for backend query results, however, the skill focuses on typed implementation which reduces typical injection surface.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 06:57 AM
Security Audit — agent-trust-hub — pbvex-client