openai-tool-calling-runtime

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface identified through the ingestion of external architectural configuration.
  • Ingestion points: The skill instructions involve loading and parsing ai-architecture.md to identify approved tools, schemas, and authorization scopes.
  • Boundary markers: There are no instructions provided for using delimiters or boundary markers to isolate content when reading the configuration file.
  • Capability inventory: The skill generates tool execution adapters, authorization checks, and audit logging hooks, which represent executable capabilities based on the ingested configuration.
  • Sanitization: The skill incorporates strong mitigations by requiring explicit authorization and input validation before any side-effecting code is executed.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 08:35 AM
Security Audit — agent-trust-hub — openai-tool-calling-runtime