Search

Every AI skill, independently scanned. Coming soon. See https://github.com/OpenLatch/saferskills.

Read-only VibeRaven guides: explain a finding, search launch guides, get a Supabase checklist.

Catch hard-coded secrets, insecure URLs, missing env vars, and unpinned packages in MCP configs.

Local-first control plane for governing AI agent runtime behavior, from preflight risk checks to live mediation and post-run audit.

Local-first observation and heuristic risk scanning for AI coding agents (Claude Code, Codex, Cursor, Copilot CLI, Gemini). Redacts secrets, flags risky tool calls, scans skills and MCP configs offline.

Reference MCP App for Elastic Security — interactive SOC dashboards inside Claude, Cursor, and other MCP hosts.

https://github.com/eren-solutions/mcp-security-audit

Security proxy for MCP servers — intercepts tool calls, detects prompt injection, masks PII, and writes audit logs

Security Scanner Ai automation via MCP. Includes scan dependencies, check headers, scan secrets. By MEOK AI Labs.

Security scanner for Model Context Protocol (MCP) servers. Detects prompt injection, tool poisoning, capability escalation, and rug-pull attacks.

Security skills MCP server for AI coding agents — secret / dependency / Dockerfile / GitHub Actions scanning, offline OSV cache, and 28 hardening skills.

MCP server for the Solana Security Standard (SOL-0XX): scan Solana/Anchor Rust and serve the rules to any MCP client (Cline, Copilot, Cursor, Claude, Windsurf).

ESLint security rules for Model Context Protocol (MCP) servers — catches SANDWORM_MODE credential harvesting, path traversal, command injection, and CVE patterns at dev time

MCP server for local security scans: Trivy for vulnerable dependencies, misconfigurations and container images, Semgrep for code patterns, Gitleaks for leaked secrets. Secret values are never returned.

it's a real agent harness with cloud-provider modules (AWS, Azure, GCP, OCI, IBM, Alibaba, PCF), a bastion security gateway, and an MCP server, plus 14 skills.

MCP server for AI agents to scan and audit smart contracts on Base for security vulnerabilities. Detect honeypots, rug pulls, hidden mints, proxy risks, and generate full audit reports.

Static security scanner for Model Context Protocol (MCP) servers. Detects tool-description poisoning, exfiltration cues, hidden-unicode payloads, arbitrary command execution, SSRF surface, hardcoded secrets, and rug-pull risk. Runs locally for Cursor and

One-command installer for Claude Code skill security audit — SKILL.md, hooks, semgrep rules, IAM templates, and updater

Zero-config security scanner for vibe-coders. 1,116 patterns, OWASP Top 10.

Secured SSH MCP server with command whitelist/blacklist filtering for safe remote server management