second-brain
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local shell script at
scripts/ensue-api.shto communicate with the Ensue backend. - [EXTERNAL_DOWNLOADS]: Uses
curlwithin theensue-api.shscript to send POST requests tohttps://api.ensue-network.ai/for retrieving and searching memories. - [DATA_EXFILTRATION]: Transmits user-provided notes, concepts, and toolbox entries to the external Ensue API for storage. This behavior is clearly documented as the primary function of the 'Second Brain' skill.
- [PROMPT_INJECTION]: Potential for indirect prompt injection (Category 8) identified:
- Ingestion points: External data enters the agent context through the
get_memoryanddiscover_memoriesoperations defined inscripts/ensue-api.sh. - Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore' instructions when presenting retrieved content to the agent.
- Capability inventory: The skill includes scripts for creating, updating, and deleting memory entries, allowing for write operations based on agent instructions.
- Sanitization: There is no evidence of sanitization or content validation performed on the data returned from the Ensue API.
Audit Metadata