api-cms-sanity

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for fetching external content from the Sanity Content Lake, which serves as a potential surface for indirect prompt injection if the ingested data contains malicious instructions.
  • Ingestion points: Data is retrieved via client.fetch as documented in SKILL.md and examples/core.md.
  • Boundary markers: The skill does not explicitly define delimiters for the ingested content within the LLM context, but it provides clear structural separation for data handling.
  • Capability inventory: The skill provides patterns for both reading data and performing mutations (create, patch, delete) in examples/mutations.md.
  • Sanitization: The skill proactively addresses security risks by mandating the use of parameterized GROQ queries ($param) and explicitly warning against string interpolation to prevent injection vulnerabilities.
  • [SAFE]: No malicious behavior, obfuscation, or unauthorized access patterns were identified. The documentation emphasizes security best practices, such as pinning API versions and using environment variables for sensitive tokens like SANITY_API_TOKEN.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 03:56 PM
Security Audit — agent-trust-hub — api-cms-sanity