skills/chann/skills/diff-viewer/Gen Agent Trust Hub

diff-viewer

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/generate_diff_report.py executes git commands using subprocess.run to retrieve diff and repository metadata. These calls use argument lists rather than shell strings, which is a secure practice that prevents command injection.- [EXTERNAL_DOWNLOADS]: The skill's HTML template loads the highlight.js library from Cloudflare's public CDN (cdnjs.cloudflare.com). This is a well-known service for delivering public library assets and is used here for legitimate syntax highlighting functionality.- [SAFE]: The skill follows security best practices by using html.escape on all content derived from the git diff before embedding it in the report, preventing potential cross-site scripting (XSS) in the generated viewer. It also explicitly instructs the agent not to edit the .gitignore file automatically, maintaining user control over the environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 08:29 AM
Security Audit — agent-trust-hub — diff-viewer