• OUTLIERS
  • PROGRESS
  • TRENDS
  • HUMAN

How to audit grant review scores with your AI

Connect Good Grants to your AI client to pull real-time scoring data. Once connected, you can query specific grant rounds to check for reviewer bias or stalled progress.

Ask AI about this page

Short answer

How can I audit grant review scores using my AI?

Your AI uses the Good Grants connector to pull review data for any specific grant round. You can then ask it to find outliers, identify slow reviewers, or summarize scoring trends. This breakdown shows how to set it up and the specific queries you can run.

Audit outcomes

What you get from an audit

Running these queries turns raw scoring data into actionable oversight.

OUTLIERS

Identify scoring bias

The AI compares individual reviewer scores against the round average. It flags anyone whose scores are significantly higher or lower than the rest of the panel.

PROGRESS

Spot review bottlenecks

By looking at completion status, your AI identifies which reviewers are lagging behind or which applications haven't been touched yet.

TRENDS

Summarize scoring patterns

The AI aggregates scores across different categories to show you where applicants are strongest or weakest across the entire round.

HUMAN

Reviewer intervention

The AI identifies specific cases that need a human to decide if a score was an error or a legitimate outlier.

The workflow

What your AI does when the data arrives.

The AI acts as a data analyst for your grant rounds by pulling specific review records.

  1. Fetch review data

    Your AI pulls the specific scoring and comment data for the grant round you specify.

    list_application_reviews

  2. Calculate averages

    It computes mean scores for each application to establish a baseline for comparison.

    list_application_reviews

  3. Detect anomalies

    The agent compares individual scores against those averages to find statistical outliers.

    list_application_reviews

  4. Report findings

    The AI presents a summary of the review progress and any flagged scoring issues.

    list_application_reviews

Try it

Copy these to start.

Use these prompts to begin your audit immediately.

Starting points

Swap out the bracketed text for your actual grant round names or IDs.

Good Grants Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_nCk4CF4yPoLGuWSzB3b0lBUumlO7tXjPJZATy0yR/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Good Grants capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "good-grants-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_nCk4CF4yPoLGuWSzB3b0lBUumlO7tXjPJZATy0yR/mcp"
    }
  }
}
Copy into chat04

  • List all reviews for the [Grant Round Name] round and show me the average score for each application.

  • In the [Grant Round Name] round, which reviewers have submitted the fewest scores so far?

  • Find any applications in [Grant Round Name] where the scores from different reviewers vary by more than 20 percent.

  • Summarize the scoring progress for the [Grant Round Name] round and tell me if we are on track to finish by the deadline.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Good Grants once, then ask.

Just click the link to connect your account. Your credentials stay encrypted, and you can start querying your data in any MCP-compatible client right away.

Connect Good Grants to your AI

FAQ

Common questions about auditing scores

  • 01

    Can the AI change a reviewer's score?

    No. The AI can only read and analyze the data provided by the connector. It cannot modify any scores or records in Good Grants.

  • 02

    Can I use this to automatically select winners?

    No. The AI provides the data and identifies patterns, but the final decision to select or reject an applicant stays with you.

  • 03

    How does the AI know which grant round to look at?

    You tell the AI which round you want to audit in your prompt, and it uses that information to filter the data it pulls.

  • 04

    Does the AI see all the reviewer comments?

    Yes, it can access the text within the reviews to help explain why certain scores were given.

  • 05

    Can the AI delete a review if it finds an error?

    No. The AI has no permission to delete or edit any data in your Good Grants account.

  • More questions about Good Grants? The Connector page answers them. See everything the Good Grants Connector can do

Connect Good Grants to Claude, Cursor, ChatGPT & more