mem-search

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes a local Python initialization script (init_db.py) and the sqlite3 command-line utility via the Bash tool to query session data stored in the ~/.softspark/ directory.
  • [PROMPT_INJECTION]: User-provided search terms from $ARGUMENTS are directly inserted into an SQLite full-text search query. While this pattern creates a surface for potential SQL injection, the skill includes explicit instructions for the agent to escape single quotes, which mitigates the risk for standard search usage.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 10:52 AM
Security Audit — agent-trust-hub — mem-search