api-smoke-validation

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFECREDENTIALS_UNSAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill consists of instructional documentation and evaluation tests for API testing workflows using the standard hurl tool. No malicious intent or instructions were found.\n- [CREDENTIALS_UNSAFE]: The skill documentation explicitly mandates secure credential management, instructing the agent to avoid hardcoding secrets in test files and instead use environment variables or parameterization (e.g., in SKILL.md under Guardrails).\n- [NO_CODE]: The skill is entirely documentation-based and does not include executable scripts, shell commands, or external dependencies that require installation or execution at runtime.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for ingesting external data (API responses) through hurl.\n
  • Ingestion points: API response bodies, headers, and status codes are read from the hurl tool output in SKILL.md (Workflow Step 2).\n
  • Boundary markers: The instructions specify using explicit assertions for status codes and response shapes, which bounds the expected input.\n
  • Capability inventory: The skill uses hurl for network requests but lacks capabilities for file-writing, subprocess execution, or direct command execution of external data.\n
  • Sanitization: The skill encourages deterministic fixtures and narrow assertions, limiting the surface for malicious payload execution through response data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 05:29 PM