Guardvibe

GuardVibe

Security infrastructure your AI can't be.

No matter how good your coding agent gets, it can't know the CVE published after its training cutoff, it can't deterministically guarantee the same check every run, it can't hold your whole repo in context, and it can't objectively review its own code. GuardVibe does all four — the deterministic, post-cutoff-current, whole-repo, author-independent verification layer for AI-written code.

• 🗓️ Knows what your AI doesn't. CVE rules refreshed daily from GHSA / OSV.dev / CISA KEV — GuardVibe flags vulnerable dependencies published *after* your model's training cutoff. (274 CVE rules, npm run intel daily triage.)
• 🎯 Deterministic, not probabilistic. Same code = same result, every run (content-hashed). Your AI guesses; GuardVibe doesn't.
• 🗺️ Sees the whole repo. Cross-file taint + auth-coverage across every route — catches the unprotected endpoint your agent's narrow context missed.
• 🔍 An independent second pair of eyes. The thing that wrote the code can't review itself. GuardVibe is the outside checker on AI-written code — in the loop *while* your AI codes (real-time edit hook), not after.
• ⬅️ NEW: Starts before the first line of code. Every scanner on earth — including your agent reviewing itself — acts *after* the code exists. secure_prompt acts *before*: it analyzes the coding prompt itself, detects the stack and attack surfaces it implies, and embeds severity-ranked GuardVibe requirements into the prompt your AI executes. The vulnerability is prevented, not caught. Deterministic, zero LLM calls — and if the prompt is already secure, it passes through untouched.

The security MCP built for vibe coding. 649 security rules, 39 tools covering the entire AI-generated code journey — from the prompt itself to production deployment.

Works with Claude Code, Cursor, Gemini CLI, Codex, VS Code (Copilot), Windsurf, and any MCP-compatible coding agent.

Why a tool, when your AI is so good?

"More rules" was never the moat — a strong model already knows most security rules by heart. What it *can't* do is be deterministic, know the CVE published after its training cutoff, hold your whole repo in context, or objectively review the code it just wrote. Those four gaps are structural; they don't close as models improve. GuardVibe is the layer that fills them — running *while* your AI codes, not in a separate audit later. And since v3.19, it runs *before* your AI codes too: secure_prompt rewrites the task itself so the security requirements are in the prompt, not in the post-mortem.

Why GuardVibe

Most security tools are built for enterprise security teams. GuardVibe is built for you — the developer using AI to build and ship web apps fast.

• 649 security rules, 39 tools purpose-built for the stacks AI agents generate
• Zero setup friction — npx guardvibe and you're scanning
• No account required — runs 100% locally, no API keys, no cloud
• Understands your stack — not generic SAST, but rules that know Next.js, Supabase, Stripe, Clerk, and the tools you actually use
• CVE version intelligence — detects 274 known vulnerable package versions in package.json, refreshed every day from GHSA / OSV.dev / CISA KEV
• AI agent & MCP security — detects MCP server vulnerabilities, tool-description prompt injection (OWASP MCP Top 10), model-controlled sandbox-disable flags, excessive AI permissions, indirect prompt injection
• Auto-fix suggestions — fix_code tool returns concrete patches and structured edits the AI agent can apply mechanically. Coverage: hardcoded credentials → env-var migration; public-prefix LLM keys (NEXT_PUBLIC_/VITE_/EXPO_PUBLIC_/REACT_APP_) → prefix removal; CORS wildcards → env allowlist; dangerouslyAllowBrowser flags → drop; sandbox bypass flags (unsafe/noSandbox/allowEval) → drop; agent loops → add maxSteps; raw-HTML React props → ; missing auth checks → insert auth guard; SQL injection → parameterized queries; missing rate limiters / CSRF / security headers → snippet templates.
• Pre-commit hook — block insecure code before it reaches your repo
• CI/CD ready — GitHub Actions workflow with SARIF upload to Security tab
• Agent-friendly output — JSON format for AI agents, Markdown for humans, SARIF for CI/CD
• Plugin system — extend with community or premium rule packs

New in v3.71.x

• @hapi/content, Socket.IO cluster engine, nodejs-whisper and obsidian-tc gaps closed — v3.71.0 adds VG1294 the @hapi/content quadratic Content-Disposition parsing DoS below 6.0.3 plus the never-fixed content 3.0.6–4.0.6 (CVE-2026-119202), VG1295 the @socket.io/cluster-engine prototype pollution via session lookup in 0.1.0 (CVE-2026-102600), VG1296 the nodejs-whisper OS command injection through audio file paths below 0.4.0 (CVE-2026-110644), VG1297 the obsidian-tc MCP server DNS-rebinding drive-by and cross-vault token access below 1.3.1 (CVE-2026-77363, CVE-2026-77366) and VG1298 its unowned note overwrite and declined-confirmation bypass in 1.3.1–1.31.3 (CVE-2026-115478, CVE-2026-115477).

New in v3.70.x

• Highcharts, decode-uri-component, follow-redirects, node-forge and joi gaps closed — v3.70.0 adds VG1289 the Highcharts chart-configuration XSS and Object.prototype pollution cluster below 11.4.9, 12.6.2 and 13.1.1 (CVE-2026-118246, CVE-2026-118251, CVE-2026-118244, CVE-2026-116796, CVE-2026-118250, CVE-2026-118249), VG1290 the decode-uri-component quadratic replacement DoS up to 0.5.0 (CVE-2026-118664), VG1291 the follow-redirects cross-realm sensitiveHeaders residual pin 1.16.1 (CVE-2026-116689), VG1292 the node-forge PEM ReDoS, PKCS#12 MAC forgery, TLS record DoS and basicConstraints bypass up to 1.4.0 (CVE-2026-118779, CVE-2026-118782, CVE-2026-119188, CVE-2026-119187, CVE-2026-119190), and VG1293 the joi email({ multiple: true }) quadratic split residual window plus the never-fixed @hapi/joi 16.x–17.x (CVE-2026-118772).

New in v3.69.x

• systeminformation, RE2JS, GraphQL Tools, Vue I18n and graphql-scalars gaps closed — v3.69.0 adds VG1284 the systeminformation Windows services() command injection via Unicode quotes residual window below 4.17.0 and 5.31.6–5.33.15 (CVE-2026-116576), VG1285 the RE2JS quadratic lookbehind matching DoS below 2.8.7 (CVE-2026-116795), VG1286 the @graphql-tools/executor-legacy-ws connectionParams replay residual pin 1.1.35 (CVE-2026-118665), VG1287 the Vue I18n escapeParameter XSS through iframe srcdoc across vue-i18n, petite-vue-i18n and @intlify/* (CVE-2026-118257), and VG1288 the graphql-scalars PhoneNumber ReDoS from 1.24.1 to 2.0.0 (CVE-2026-115887).

New in v3.68.x

• MCP SDK adapters, TinyMCE, dd-trace, CKEditor 5 and Seroval gaps closed — v3.68.0 adds VG1279 the MCP TypeScript SDK Node / Hono adapter unbounded request body and JSON-RPC batch residual window (GHSA-f897-62hf-6gxq), VG1280 the TinyMCE media plugin data-mce-object stored XSS on 7.9.3 and 8.5.1–8.9.2 (CVE-2026-116675), VG1281 the dd-trace query-string obfuscation ReDoS below 5.131.1 / 6.20.1 (CVE-2026-116693), VG1282 the CKEditor 5 engine XSS through es-toolkit prototype pollution below 47.7.4 / 48.5.1 (CVE-2026-116688), and VG1283 the Seroval fromJSON array-length / sequence-index DoS and async-iterator callable invocation residual window 1.6.3–1.6.7 (CVE-2026-116581 / CVE-2026-116584 / CVE-2026-116580).

New in v3.67.x

• git-diff, happy-dom, postal-mime, Electron and undici gaps closed — v3.67.0 adds VG1274 the git-diff OS command injection through gitDiff() below 2.0.7 (CVE-2026-110634), VG1275 the happy-dom ESM import-alias and export-name code injection plus iframe / stylesheet SSRF below 21.0.4 (CVE-2026-115895 / CVE-2026-33943 / CVE-2026-118660), VG1276 the postal-mime quadratic RFC 2047 decoding DoS (GHSA-f2hq-32vh-53cc), VG1277 the Electron compromised eMudhra intermediates residual window (CVE-2026-115570) and VG1278 the undici Agent TLS callback drop and TLS 1.2 renegotiation crash residual window on 5.x / 6.11 (CVE-2026-109383 / CVE-2026-87018).

New in v3.66.x

• zod, undici, body-parser, multer, GraphQL Tools and MSAL Node gaps closed — v3.66.0 adds VG1268 the zod unbounded array issue accumulation memory exhaustion below 4.6.6 (CVE-2023-54404), VG1269 the @graphql-tools/utils mergeIncrementalResult prototype pollution residual pin (CVE-2026-114604), VG1270 the undici excessive content-encodings process crash residual window (CVE-2026-109869), VG1271 the body-parser trailing-bytes decoder crash (CVE-2026-111181), VG1272 the multer pending async fileFilter crash residual window (CVE-2026-113872) and VG1273 the @azure/msal-node auth code flow parameter disclosure (CVE-2026-109461). 249 CVE version-pin rules.

New in v3.65.x

• Next.js, DOMPurify, Sentry and toml gaps closed — v3.65.0 adds VG1263 the Next.js residual window for the Image Optimization RCE, the use cache typed-array RCE and the Server Actions DoS (GHSA-2464-fhr4-w9xw, GHSA-673q-mjm6-r5xr, CVE-2026-94482), VG1264 and VG1265 the DOMPurify SAFE_FOR_TEMPLATES shadow-root bypass and IN_PLACE hook-detached subtree XSS (GHSA-cg2g-xh64-v4q7, GHSA-mrv2-3pqr-j5jr), VG1266 the @sentry/nextjs middleware bypass under the tunnel route (CVE-2026-114139) and VG1267 the toml quadratic and unbounded-integer DoS cluster (CVE-2026-114141 / 114143 / 114144). 243 CVE version-pin rules.

New in v3.64.x

• Storybook and ApostropheCMS gaps closed — v3.64.0 adds VG1261 the Storybook dev-server code injection from a malicious website via missing origin validation (CVE-2026-114125, critical) across storybook 8.x–10.x and the @storybook/core, core-server, experimental-addon-test, addon-vitest and addon-mcp packages, and VG1262 the ApostropheCMS denial-of-service cluster — pre-auth import uploads, leaked CSV temp files and prototype-member deletion (CVE-2026-113711 / 113658 / 113666). 238 CVE version-pin rules.

New in v3.63.x

• MCP SDK, follow-redirects, markdown-it, Vite and Angular router gaps closed — v3.63.0 adds VG1256 the MCP TypeScript SDK HTTP-server memory exhaustion, cross-session task access and cross-client message delivery (CVE-2026-111768 / 111767 / 111770), VG1257 the follow-redirects credential-header leak on cross-origin redirects (CVE-2026-111756 / 111755), VG1258 the markdown-it quadratic-time rendering DoS pair (CVE-2026-111758 / 111759), VG1259 the Vite / launch-editor dev-server command injection on Windows (CVE-2026-112942 / 112943) and VG1260 the Angular router SSR DoS pair (CVE-2026-111751 / 111752). 236 CVE version-pin rules.

New in v3.62.x

• Qwik and TinaCMS CLI gaps closed — v3.62.0 adds VG1254 the Qwik server crash from a single unauthenticated request carrying a resource thenable in the qwik-json body (CVE-2026-84994) and VG1255 the TinaCMS CLI code injection through an unescaped Git branch name in the generated client (CVE-2026-108259). 231 CVE version-pin rules.

New in v3.61.x

• TinaCMS, JHipster, PraisonAI, Actual and Backstage gaps closed — v3.61.0 adds VG1249 the Backstage Sentry Scaffolder actions SSRF and integration token disclosure (CVE-2026-106459), VG1250 the Actual Sync Server CORS proxy GitHub