memory
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a local Python script (bin/memory-compiler.py) to process and compile memory entries stored in the user's vault. This is a standard operation for a utility-based skill.- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes data from user-controlled directories (vault and 00-Inbox/).
- Ingestion points: Reads from the 00-Inbox/ directory and general vault notes.
- Boundary markers: The instructions do not define specific delimiters to separate untrusted data from agent instructions.
- Capability inventory: Accesses local file system and executes a Python script for data processing.
- Sanitization: No explicit sanitization or validation steps for incoming data are described.
Audit Metadata