api-idempotency-retry

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze external, untrusted data such as endpoint lists, OpenAPI specifications, and business logic descriptions to identify retry-unsafe operations and design recovery points. This creates a surface for indirect prompt injection where malicious instructions could be embedded within the technical documentation to manipulate the agent's output.
  • Ingestion points: Step 1 ('Identify retry-unsafe endpoints') and the 'Clarifying questions' section require the agent to process user-provided endpoint details and architecture specifications.
  • Boundary markers: The instructions lack explicit boundary delimiters or warnings to ignore instructions embedded within the processed technical data.
  • Capability inventory: The skill is primarily descriptive and does not utilize high-risk tools for file system modification, network exfiltration, or shell execution.
  • Sanitization: No validation or sanitization routines are specified for the agent to apply to the technical specifications it analyzes.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:17 PM
Security Audit — agent-trust-hub — api-idempotency-retry