claude-stats
Pass
Audited by Gen Agent Trust Hub on Jul 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill accesses Claude Code session transcripts stored in
~/.claude/projects/. These files contain sensitive interaction history, but the access is limited to reading the data for local aggregation. No evidence of data exfiltration or unauthorized transmission was found. - [SAFE]: The skill executes a local Python script to perform data analysis. The script uses a standard argument parser to handle user inputs for the period and report type, ensuring that the script only processes expected commands.
Audit Metadata