spark-memory
Pass
Audited by Gen Agent Trust Hub on May 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates interaction with a specialized CLI tool named
spark. This tool is used to perform searches and submit insights to a shared memory store. All execution patterns are consistent with the skill's stated purpose of documentation and knowledge sharing. - [DATA_EXFILTRATION]: The skill is designed to transmit development insights, architecture patterns, and best practices to an external shared memory service. This data transmission is the primary function of the skill. The instructions provide explicit safety guidelines, directing the agent never to share secrets, API keys, or other sensitive information.
- [PROMPT_INJECTION]: As the skill processes and acts upon data retrieved from a shared team memory (via
spark query), it has a surface for indirect prompt injection. However, this is a known characteristic of collaborative memory tools. The skill uses structured XML formatting for query responses, which helps in differentiating between data and instructions.
Audit Metadata