Works with the AI Agents you already use
web app security checklist
A 20-point security audit skill for web apps, focusing on secrets, access control, AI abuse, and payment safety.
Free
web app security checklist
Example session with this skill installed
I'm about to launch a Next.js app with Supabase and Stripe. Can you run a security check on my current repo to see if it's safe to ship?
- Read your context and instructions
- Compiled the web app security
I have completed the security audit for your Next.js, Supabase, and Stripe integration. I found one critical leak and two high-priority configuration issues.
Security Review: Project X Launch
Summary: The application has a critical secret leak in the frontend bundle and missing RLS policies on two Supabase tables. Payment logic is correctly verified via webhooks.
Critical
[1] Leaked Supabase Service Role Key | where: src/lib/supabase.ts | fix: Replace SERVICE_ROLE_KEY with NEXT_PUBLIC_SUPABASE_ANON_KEY for client-side use and move admin logic to a Server Action.
High / Medium / Low
[8] Missing RLS Policies | High | public.user_profiles and public.orders have RLS enabled but no policies, blocking all non-admin reads.
[11] No Login Rate Limiting | High | The /api/auth route lacks a rate limiter, leaving it open to brute-force attempts.
[18] Security Headers | Medium | Missing Content-Security-Policy and Strict-Transport-Security headers in next.config.js.
Results (Top 5)
| # | Item | Status | Notes |
|---|---|---|---|
| 1 | Frontend Secrets | FAIL | Admin key exposed in browser bundle. |
| 8 | DB RLS | PARTIAL | Tables locked but no access policies defined. |
| 17 | Payment Verif | PASS | Stripe webhook signatures verified correctly. |
| 2 | .env in Git | PASS | .env correctly listed in .gitignore. |
| 7 | IDOR Checks | PASS | owner_id checks present in all Supabase queries. |
Next steps
- Rotate the Supabase Service Role Key immediately in the dashboard as it was exposed to the browser.
- Apply RLS policies to allow users to read their own data using
auth.uid(). - Add
upstash/ratelimitor similar middleware to your Next.js API routes.**
Connects securely to your tools. The creator never sees your data.
About this skill
Ship web applications with confidence by identifying vulnerabilities before they reach production. This skill provides a systematic 20-point security audit tailored for modern web stacks, including specific checks for AI-integrated features and payment flows. It moves beyond generic advice to provide actionable tests and minimal fixes for real-world codebases.
What it does
- Secret Detection scans for leaked API keys in frontend bundles, environment files, and git history.
- Access Control Audit verifies server-side permissions, IDOR vulnerabilities, and database row-level security.
- Abuse Prevention checks rate limiting for auth endpoints and sets usage caps for AI API calls.
- Input Validation identifies injection risks and cross-site scripting (XSS) in user-generated content.
- Payment Verification audits webhook signatures and server-side fulfillment logic.
How it works
- Application Mapping where you provide the tech stack, auth methods, and high-risk features like payments or AI.
- Code Inspection where the agent analyzes your repo, configuration files, and API handlers.
- Vulnerability Scoring which marks items as PASS, FAIL, or PARTIAL based on severity levels.
- Remediation Reporting delivering a prioritized list of fixes and a formal security summary.
Frameworks & tools
Works with Next.js, React, Supabase, Firebase, Node.js, Django, and Laravel. Supports integration with scanners like gitleaks and trufflehog.
Why this beats prompting it yourself
Generic prompts often miss the nuances of client-side vs server-side execution in modern frameworks. This skill enforces a strict "least privilege" mindset and specific testing procedures, like cross-account IDOR checks, that generic LLMs typically overlook.
Use cases
- Pre-launch Audit to catch leaked keys or missing rate limits before going live.
- PR Security Review for checking new payment or file upload logic.
- AI Cost Control to ensure user-facing LLM features are capped and protected from prompt injection.
- Environment Hardening to verify production configs, CORS policies, and security headers.
Known limitations
Does not perform active penetration testing or dynamic scanning of live production environments. Requires manual permission for destructive actions like rewriting git history.
Cannot perform active penetration testing or live traffic interception. Manual confirmation required for git history rewrites or key revocations. Focuses on code and config, not infra-level OS hardening.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
2 uses
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account