secret-scanning

To scan staged git diffs for exposed secrets, provide the repository owner and name or file content inputs. The skill integrates with the GitHub MCP run_secret_scanning tool to produce location-tagged findings and confidence indicators for staged changes.

Yes, you can scan single content snippets for credentials. The skill processes pasted text or specific repository files to identify likely secrets and returns their locations and confidence levels.

Secret scanning with GitHub MCP uses the run_secret_scanning tool to perform server-side detection of exposed credentials. It analyzes repository content, file inputs, or git diffs and returns structured findings with location tags and confidence indicators.

You need to provide the repository owner and name or file content as inputs to detect secrets in committed files. The skill uses these inputs to run the GitHub MCP secret scanning tool and identify potential credential leaks in repository content.

When scanning local files for secrets, avoid scanning large generated directories by default. The skill includes safety practices that require user confirmation before scanning ignored or sensitive local files to prevent unintended exposure or performance issues.

Yes, this approach catches high-entropy values in configuration files. The skill identifies exposed secrets, credentials, and high-entropy values in files, code snippets, and git diffs to reduce the risk of accidental leaks and deployment failures.