The Agent Skills Directory

[SAFE]: The skill implements robust security measures for processing untrusted data (git diffs). The render-diff.awk script contains a dedicated rd_html function that escapes HTML special characters (`, <, >, ", ') to prevent Cross-Site Scripting (XSS) in the generated reports.
[SAFE]: File system operations are handled securely. The rd_prepare_workspace function in scripts/lib/filesystem.sh uses mktemp -d and sets a restrictive umask 077 to ensure temporary files are not accessible by other users.
[SAFE]: The shell scripts follow defensive programming practices, including set -euo pipefail and careful validation of input/output paths to prevent accidental file overwrites or unauthorized access.
[SAFE]: The SKILL.md instructions explicitly warn the AI agent against using eval or executing the contents of supplied diffs, mitigating potential command injection or remote code execution risks.
[SAFE]: No external network calls, hardcoded credentials, or obfuscated code patterns were detected.

render-reading-diff