analyze-usage

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes local command-line tools including duckdb, find, stat, and jq to aggregate data and manage a persistent database. This is a standard operation for the tool's analytical purpose.
  • [PROMPT_INJECTION]: The skill ingests data from external logs which may contain untrusted content, creating an indirect prompt injection surface. Evidence Chain: (1) Ingestion points: Claude Code JSONL logs and Cursor SQLite databases. (2) Boundary markers: None specific. (3) Capability inventory: Local file read/write and SQL execution. (4) Sanitization: Escaping of single quotes and structured JSON extraction.
  • [SAFE]: The tool operates entirely on local data and does not perform network operations. Sensitive file access to chat logs is restricted to the tool's primary stated purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:57 AM