/scan-input

Scans input for specified patterns or issues and returns a detailed report of findings.

Runs an agentic security scanner on the project, detecting vulnerabilities across SCA, secrets, authz, MCP, pipeline, and logic surfaces. Supports focused scans per surface and supplementary output blocks.

Orchestrates security scanners (gitleaks, semgrep, etc.) with AI-powered triage to reduce false positives, explain exploitability, and suggest fixes. Analyzes git diffs or specified files.

Scans skills, plugins, MCP tools, and code for prompt injection, PII harvesting, credential theft, and supply chain attacks. Supports --deep and --report flags.

Runs a fast security and quality scan on specified files or directories, checking for hardcoded secrets, injection flaws, debug leftovers, and code quality issues. Produces a categorized report.

Runs a full static security audit on a file or directory, reporting exploitable vulnerabilities with severity, CWE, and remediation snippets.