The Agent Skills Directory

[SAFE]: The skill is focused on security best practices, specifically helping users redact sensitive information from evidence before it is shared. It promotes data privacy and responsible disclosure.
[COMMAND_EXECUTION]: The skill includes a shell function, sanitize_har, which uses the standard command-line tool jq to strip sensitive headers (like Cookies and Authorization) from HAR files. The script is provided in plain text, is easy to audit, and performs a strictly defensive task.
[DATA_EXPOSURE]: Rather than exposing data, this skill is specifically designed to prevent the accidental exposure of session tokens, API keys, and personal information during the bug reporting process.

evidence-hygiene