api-cms-strapi

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill documentation and examples follow security best practices for headless CMS development. No malicious patterns, obfuscation, or unauthorized access attempts were detected.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for processing data from a Strapi CMS backend, which acts as an untrusted external data source. While this creates a potential surface for indirect injection, the skill proactively mitigates this by mandating the use of Strapi's built-in sanitization and validation helpers.
  • Ingestion points: Data returned from Document Service API calls and REST API responses in examples/core.md and examples/backend.md.
  • Boundary markers: Not explicitly defined for prompt instructions, but the skill emphasizes strict data-level boundaries using sanitization methods.
  • Capability inventory: Includes CMS CRUD operations via strapi.documents() and external network requests via fetch for authentication and data retrieval.
  • Sanitization: Strongly enforces the use of sanitizeQuery, sanitizeOutput, and validateQuery in all custom controllers to prevent role-based data exposure and filter malicious inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 03:56 PM
Security Audit — agent-trust-hub — api-cms-strapi