ddd4j-cache
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive instructional content for choosing and configuring cache providers within the ddd4j ecosystem, focusing on implementation correctness and security.
- [DATA_EXFILTRATION]: Instructions include specific guidance to prevent the exposure of sensitive data by excluding PII from cache keys and logs, recommending internal IDs or hashing as alternatives.
- [DYNAMIC_EXECUTION]: The skill contains defensive security advice, specifically warning against insecure deserialization configurations (e.g., DefaultTyping) that could be exploited if applied to untrusted inputs.
- [INDIRECT_PROMPT_INJECTION]: The skill defines boundaries for data ingestion. 1. Ingestion points: Cache data retrieved from Redis or local stores. 2. Boundary markers: Explicit instructions to use trusted data boundaries for serialization. 3. Capability inventory: Advice provided for implementation of atomic contracts and locking. 4. Sanitization: Guidance to exclude PII and implement strict invalidation policies.
Audit Metadata