conversation-analyzer

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill reads the user's conversation history from ~/.claude/history.jsonl. This is the core functionality of the skill, and the data is processed locally without external transmission.
  • [COMMAND_EXECUTION]: The skill uses the bash tool to run a local Python script (scripts/analyze_history.py). This script uses standard libraries (json, re, pathlib) to perform heuristic analysis on the history file.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from previous conversations. However, since the analysis is performed via hardcoded Python logic (regex and keyword counting) rather than dynamic LLM interpretation of the raw history, the risk is minimal.
  • Ingestion points: Reads ~/.claude/history.jsonl in scripts/analyze_history.py.
  • Boundary markers: Not applicable as analysis is heuristic-based.
  • Capability inventory: Includes file writing (analysis reports) and bash execution (invoking the analysis script).
  • Sanitization: Employs standard JSON parsing and regular expressions to extract specific patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 05:03 AM
Security Audit — agent-trust-hub — conversation-analyzer