remember
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to calculate project-specific directory paths and search for existing entries within the memory file. Evidence: usage of
pwd,sed, andgrepin the workflow steps. - [PROMPT_INJECTION]: The skill creates a surface for Indirect Prompt Injection by storing user-provided content in a persistent file (
MEMORY.md) that is loaded into the agent's context in subsequent sessions. Ingestion points: user input processed via/si:remember. Boundary markers: markdown list formatting. Capability inventory: file write operations and shell execution. Sanitization: instructions for the agent to parse and condense input before storage.
Audit Metadata