mem-learn

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the npx command to run the @codevoyant/agent-kit utility for indexing and searching stored knowledge. This tool is associated with the skill's author.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it stores arbitrary text provided by users and subsequently processes that content to answer queries.
  • Ingestion points: Markdown files created in the styleguide/ and recipes/ directories based on user input during Learn Mode.
  • Boundary markers: No delimiters or specific safety instructions are present to prevent the agent from following instructions embedded in stored files.
  • Capability inventory: File system read and write operations, and command execution via the npx tool.
  • Sanitization: The skill does not perform content validation or sanitization before storage or retrieval.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 11:51 PM
Security Audit — agent-trust-hub — mem-learn