Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    shark tank

    1

    Simulate a high-stakes Shark Tank episode to stress-test startup ideas and generate fundability reports.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    shark tank

    shark tank

    Example session with this skill installed

    /shark-tank ./my-saas-project --panel usa --ask "$100k for 5%" --mode manual

    • Read your context and instructions
    • Compiled the shark tank

    I have analyzed the repository for my-saas-project and seated the USA panel: Kevin (The Valuation Hawk), Mark (The Scale King), Lori (The Product Queen), Daymond (The Branding Pro), and Barbara (The Culture Expert).

    Watch it live: http://localhost:4740

    Shark Tank Episode: my-saas-project

    The founder has delivered the pitch. The sharks are now circling.

    SharkStatusFeedback
    KevinOUT"The valuation is purely speculative. You're asking for a $2M post-money on a project with no MRR."
    LoriIN"I see a hero product here. I'll offer $100k for 15%, contingent on a patent filing."
    MarkIN"I like the tech stack, but you're a developer, not a CEO yet. $100k for 20% and I'll bridge the sales gap."
    DaymondOUT"Too early for me to put my brand behind it. I'm out."
    BarbaraOUT"I don't trust the team's ability to pivot when the market shifts."

    Current Offers

    • Lori: $100,000 for 15% equity (Requires patent)
    • Mark: $100,000 for 20% equity (No contingencies)

    Negotiation Phase (Round 1/2)
    The sharks are waiting for your response. You asked for $100k at 5% ($2M valuation). Lori is at $666k valuation; Mark is at $500k.

    Next steps

    1. Respond to Lori or Mark in chat to accept, counter, or walk away.
    2. Check the Pitch Doctor section in outcome.html for the valuation fix.
    3. Open the qa.json file to see the private gut-reads from the sharks.

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

    About this skill

    Getting a project funded requires more than just a good idea, it requires surviving the scrutiny of veteran investors. This skill subjects your project, repo, or pitch deck to a simulated high-stakes investment panel modeled after Shark Tank India and USA styles.

    What it does

    • Multi-Agent Simulation orchestrates a founder agent and a panel of sub-agent sharks with researched investment personalities.
    • Dynamic Pitch Generation extracts data from your code or README to build a 6-8 slide deck following professional frameworks.
    • Aggressive Q&A subjects your business model to two rounds of grilling on valuation, traction, and market fit.
    • Automated Negotiation handles counter-offers and equity splits based on your pre-defined walk-away terms.
    • Live Monitoring streams the episode to a local Node.js web room for real-time viewing of slides and banter.
    • Pitch Doctoring generates a comprehensive HTML report with fixed pitches, objection banks, and valuation-saving advice.

    How it works

    1. Initialize: Point the agent at a repo, URL, or text description and specify your panel (India, USA, or Global) and investment ask.
    2. Pitch: The founder agent generates slides and an elevator pitch while you watch the live room at localhost:4740.
    3. Defend: In Auto mode, the agent negotiates; in Manual mode, you answer shark questions and decide on offers in the chat or web UI.
    4. Close: The panel delivers scores, feedback, and final investment verdicts (In or Out).

    Frameworks & tools

    Built for Node.js 20+ environments. It utilizes sub-agent orchestration (Claude Code/Cursor), JSON-based state management, and a custom HTML/CSS report builder for self-contained portability.

    Why this beats prompting it yourself

    Single prompts lack the adversarial tension and specialized personas required for a realistic roast. This skill uses a multi-round state machine where sharks react to each other and your specific data points, preventing the "hallucination of agreement" common in standard LLM chats.

    Use cases

    • Auditing a startup idea for valuation sanity before meeting real investors.
    • Stress-testing a project README to see where the business logic is thin.
    • Practicing high-pressure negotiation and objection handling in Manual mode.

    Known limitations

    Requires a Node 20+ runtime for the live room and report builder. Founder traction data must be provided as it will not invent metrics.

    Requires Node 20+ for reports/live room. Does not invent traction metrics. Sub-agent spawning works best in Claude Code or Cursor.

    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

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 6 days ago

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files
    Listed6 days ago

    What's inside

    Frequently Asked Questions

    This skill uses a multi-agent state machine to create adversarial tension, ensuring 'sharks' challenge your data and each other. Unlike a single prompt, this prevents the AI from being overly agreeable, providing a realistic 'roast' of your business logic.

    The skill is designed for Node.js 20+ environments and is optimized for advanced orchestrators like Claude Code or Cursor. It includes a local server that streams the simulation to a web UI at localhost:4740 for a real-time visual experience.

    In Manual mode, you personally answer the sharks' questions and negotiate equity splits through the chat or web UI. In Auto mode, the founder agent uses your pre-defined 'walk-away' terms to negotiate the deal on your behalf.

    You receive the full skill logic, the local Node.js web room for live monitoring, and the custom HTML/CSS report builder. It also includes pre-configured investment personalities for India, USA, and Global panels.

    The skill generates a 'Pitch Doctor' HTML report containing a fundability score, a bank of anticipated objections, and specific advice on how to improve your valuation and business model.

    The agent extracts information directly from your provided repository, URL, or text description. While it analyzes existing data, it will not invent traction metrics, ensuring the feedback is based strictly on your project's current reality.

    New in Agents & Orchestration