More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    AI App Readiness Coach - Get your App Ready for Real Users

    3

    A 10-point technical audit and coaching framework to verify if AI-generated apps are safe and ready for launch.

    Free

    1 useSecurity scanned
    AI App Readiness Coach - Get your App Ready for Real Users

    AI App Readiness Coach - Get your App Ready for Real Users

    Example session with this skill installed

    I just built a CRM with Lovable and I'm worried about security. Can you check if it's ready for me to invite my first 50 clients?

    • Read your context and instructions
    • Compiled the ai app readiness

    I will guide you through ten readiness checks for your Lovable app. I have access to web search to check current Lovable documentation and will use a browser to inspect your public pages. I cannot see your private data or code unless you provide it.

    We will start with the two most critical checks regarding data exposure.

    1. Who can see your data?

    This is to ensure one client cannot see another client's records.

    • Test 1 (Logged-out): Open your dashboard link in an Incognito window. You should see a login page, not data.

    Test 2 (Two-account): Create two test accounts (TEST-A and TEST-B). Try to access a TEST-A record link while logged in as TEST-B.

    • Builder Question: Ask your builder: "Is Row Level Security (RLS) enabled on all Supabase tables?"

    Score: Yellow (Pending your test results)

    2. Is a secret showing?

    I have scanned your public scripts at the URL provided.

    • Result: I found a string starting with pk_live_. This is your Stripe Public Key, which is safe to be public.
    • Action: Search your source code for service_role or sk_live. Do not paste them here if found.

    Score: Green

    Current Readiness Scorecard

    CheckStatusNote
    1. Data Visibility🟡 YellowAwaiting two-account test results.
    2. Secret Exposure🟢 GreenNo master keys found in public scripts.
    3. Account Ownership⚪ GrayNot started.
    4. Data Portability⚪ GrayNot started.

    Next steps

    1. Run the "Two-account test" described in Check 1 and tell me if Account B could see Account A's data.
    2. Confirm if you have the login credentials for the GitHub repository and the Supabase dashboard.
    3. Tell me the email addresses used for your Stripe and Domain accounts.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    You have an app built with Lovable, Bolt, v0, or another AI builder, but you aren't sure if it is actually safe to launch. Most AI-generated apps suffer from hidden data leaks, hardcoded secrets, or brittle permissions that non-technical owners miss. This skill provides a structured, developer-led coaching session to audit your application before real users touch it.

    What it does

    • Data security audit guided checks to ensure row-level security is active and users cannot access each other's data.
    • Secret detection scans public code and scripts for exposed Stripe, Supabase, or OpenAI API keys.
    • Ownership verification confirms the owner controls the domain, repository, and billing accounts rather than the freelancer or AI tool.
    • Functional validation tests payment paths, form deliveries, and automated jobs without risking real transactions.
    • Compliance and accessibility checks for privacy policy alignment and basic WCAG keyboard navigation.

    How it works

    1. Identify the stack you provide the AI builder used and the app's public URL to set the audit context.
    2. Execute ten checks the agent guides you through one test at a time, scoring each as Green, Yellow, or Red.
    3. Generate a scorecard the agent compiles a technical table of findings and a launch-readiness assessment.
    4. Draft builder comms it produces a specific list of technical questions or urgent fix requests to send to your developer.

    Frameworks & tools

    Optimized for apps built with Lovable, Base44, Bolt, Replit, v0, and Supabase. It uses browser-based testing patterns and platform-specific documentation.

    Why this beats prompting it yourself

    Generic prompts often miss the specific "gotchas" of AI builders, like unpublishing a site without pausing the database. This skill enforces a strict safety protocol, including specific search strings for leaked secrets and a "two-account" test methodology that reveals broken permissions.

    Use cases

    • Auditing a freelancer's work before final payment.
    • Checking a "v0" or "Lovable" prototype before moving it to production.
    • Responding to a suspected data leak by following the emergency closure steps.

    Known limitations

    Cannot bypass logins or view private databases directly. Requires the owner to perform manual browser actions if no tool access is provided.

    Cannot bypass logins or view private databases. Relies on owner-reported results for behind-the-login checks. Not a substitute for legal or professional security advice.

    How to use

    1. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 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

    1 use

    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

    Permissions

    Network Access
    Browser

    If you want it to check any hosted app or currently published project, it needs to be able to see it on the internet.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This coach is specifically designed to catch common pitfalls unique to AI-generated code, such as hardcoded API keys and broken Row-Level Security (RLS) in Supabase, which standard security tools often overlook. It uses a structured 10-point audit to ensure your prototype is safe for real users, not just functional.

    The skill is optimized for the modern AI-builder ecosystem, including Lovable, Bolt, v0, Replit, and Supabase. While it works best with these stacks, the underlying security and ownership principles can be applied to almost any web application built with a modern frontend and backend.

    No technical background is required; the agent guides you through a 'two-account' test and provides specific search strings to find vulnerabilities. If issues are found, it generates a clear, technical punch-list that you can send directly to your developer or freelancer to fix.

    You receive the full 10-point technical audit framework, the launch-readiness scorecard generator, and a communication template for coordinating fixes with developers. The purchase also includes access to the emergency closure protocol in case a leak is detected during testing.

    For security reasons, the agent does not require direct access to your database or private code repository. It functions as a guided coach, telling you exactly what to look for and where to click in your own dashboard to verify your app's safety.

    New in Agents & Orchestration