session-report

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • Local Transcript Analysis: The skill executes a Node.js script (analyze-sessions.mjs) to read usage data from ~/.claude/projects. This directory contains sensitive session history, including prompts and code. The analysis is used solely to populate a local HTML report.
  • Local File Operations: The skill writes a temporary JSON file to /tmp/session-report.json and generates the final HTML report in the current working directory. These operations are restricted to the local file system.
  • Command Execution: The skill uses node to run the bundled analysis script and cp to prepare the HTML template. This is consistent with the skill's documented purpose of generating a report.
  • Data Handling: The HTML template includes a basic escaping function (esc) to sanitize text snippets from the transcripts before they are rendered in the browser, which helps mitigate potential cross-site scripting (XSS) considerations when viewing the report.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 05:20 PM