# Grade Startup Applications with One MCP. AI Agent Connect > Accelerator Application Quality Score provides a diagnostic engine to evaluate startup pitches. It uses a specialized scoring algorithm to assess core business metrics, including problem clarity, solution uniqueness, team completeness, traction evidence, and market size validation. With this MCP, you can generate detailed reports, pinpoint weak areas, and get strategic recommendations. You can use the `calculate_application_score` tool to assess a single pitch, check performance thresholds with `get_dimension_benchmarks`, or rank multiple startups using `compare_applications` to help with shortlisting. ## Overview - **Category:** business-intelligence - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_jzP4NK1jm33dd0NP6aFfcuMCufj9juqr2jZ0gJ25/ai-agent-connect - **Tags:** startup, accelerator, scoring, evaluation, business-metrics ## Description Need to grade a batch of startup pitches? This MCP gives you a diagnostic engine for evaluating accelerator applications. It runs a specialized scoring algorithm across key business metrics: problem clarity, solution uniqueness, team completeness, traction evidence, and market size validation. Instead of gut feeling, you get hard data. You can run reports that give a quality score, flag weak areas, and suggest concrete ways to improve. Need to compare two pitches? You can rank them side-by-side. The result is a clear, actionable report that helps you make fast, informed decisions about which startups to move forward with. ## Tools ### compare_applications Use this tool to evaluate the relative quality of multiple applications. It helps you rank or shortlist startups based on their scores. ### calculate_application_score This tool provides a quantitative and qualitative assessment of a single accelerator application. You get a full report, a quality score, and specific improvement recommendations. ### get_dimension_benchmarks Check performance thresholds with this tool. It retrieves the specific scores that define 'weak' versus 'strong' performance for every dimension. ## Prompt Examples **Prompt:** ``` Evaluate this startup: Problem clarity is 8, solution uniqueness is 7, team completeness is 90%, traction is 5, market size is 8, and the application is complete. ``` **Response:** ``` The startup has a quality score of 76/100. The main weak area is traction evidence. To improve, the startup should focus on providing more quantifiable proof of market interest, such as user growth or revenue. ``` **Prompt:** ``` Compare these two applications: App A has a score of 85 and App B has a score of 72. ``` **Response:** ``` App A is ranked 1st and App B is ranked 2nd. The average score is 78.5 with a score spread of 13. ``` **Prompt:** ``` What is the threshold for a strong performance in traction? ``` **Response:** ``` The threshold for traction is a score of 7 or higher to avoid being flagged as a weak area. ``` ## Capabilities ### Score Single Pitches Run `calculate_application_score` when you need a detailed, quantitative assessment of one startup. ### Rank Multiple Startups Use `compare_applications` when you need to quickly shortlist and compare several pitches against each other. ### Check Performance Baselines Call `get_dimension_benchmarks` to understand what scores truly define 'weak' versus 'strong' performance. ## Use Cases ### Initial Deal Flow Screening You receive 50 applications in a week. Use the MCP to score them all and immediately filter out the low-potential pitches. ### Pitch Deck Feedback A founder asks, 'How good is our pitch?' Use the MCP to score their application and generate a report with actionable recommendations. ### Portfolio Review You need to compare two companies in your portfolio. Use the MCP to rank them and see which one has a stronger overall profile. ### Setting Internal Standards Before grading, you use the MCP to check the benchmarks, ensuring your team is grading against the same objective standards. ## Benefits - It provides a clear, quantifiable score instead of relying on subjective gut feelings. - You get specific feedback on weak areas, telling you exactly where a startup needs to improve. - It allows you to rank multiple companies quickly, streamlining your shortlisting process. - You can check the underlying performance thresholds, so you know what 'good' really means. ## How It Works Connecting this MCP lets your AI client run a specialized diagnostic engine against startup data. You simply tell your agent what you want to evaluate, and it returns a structured, scored report. 1. Connect your preferred AI client (Claude, Cursor, etc.) to the Vinkius catalog. 2. Tell your agent to run a score using `calculate_application_score` and provide the necessary metrics. 3. The MCP processes the data, generating a quality score and identifying weak areas. 4. You can then use `compare_applications` to rank the results or `get_dimension_benchmarks` to check the scoring rules. ## Frequently Asked Questions **What metrics does this MCP use to score a startup?** The MCP assesses core business metrics, including problem clarity, solution uniqueness, team completeness, traction evidence, and market size validation. These metrics form the basis of the quality score. **Can I use this to compare multiple companies?** Yes. You use the `compare_applications` tool to evaluate multiple pitches. It ranks them and provides an average score and score spread for comparison. **Is this tool just a simple scoring system?** No. It's a diagnostic engine. After giving a score, it provides a full report, flags specific weak areas, and offers strategic recommendations for improvement. **What does 'traction evidence' mean in this scoring?** Traction evidence refers to quantifiable proof of market interest. The MCP uses this metric to assess if the startup has concrete proof of user growth or revenue. **Do I need to know the scoring rules beforehand?** You don't. You can use the `get_dimension_benchmarks` tool to retrieve the specific performance thresholds, which tells you what scores are considered 'weak' versus 'strong'.