What problem does it solve? Leaked credentials posted in GitHub issues, PRs, comments, and discussions expose the project to compromise, and cleaning them up safely requires careful redaction, edit-history purging, author notification, and alert resolution without ever re-exposing the secret. ## Core Features & Use Cases - Alert Triage and Content Fetching: List open secret scanning alerts and fetch affected content with hide_secret=true so plaintext secrets never appear in terminal output. - Location-Aware Remediation: Redact issue/PR bodies in place, delete and recreate leaked comments (including discussion comments via GraphQL), and notify authors with the correct template per location type. - Safe Resolution and Reporting: Resolve alerts with maintainer-only resolution comments and produce a verbatim summary table of actions taken and remaining edit-history risks. - Use Case: A maintainer receives a GitHub Secret Scanning alert for a Discord bot token posted in an issue comment; the skill fetches the content, deletes and recreates the comment in redacted form, notifies the author to rotate the credential, and resolves the alert. ## Quick Start Ask the agent to list all open secret scanning alerts for the repository and process alert number 72 end to end.