- PASS/FAIL
- LOGS
- DIFFS
- HUMAN
Short answer
How can I verify my AI's code changes?
Your agent uses the verify_delivery tool to run test suites and pull the resulting logs. You get actual execution data to confirm the code works as intended. This gives you a clear way to check the work.
The outcomes.
Where your verification lands.
The AI processes the code and delivers these specific results.
PASS/FAIL
Test execution status
The AI runs your tests and tells you exactly which ones passed or failed. You get the direct result of the run.
LOGS
Raw execution data
The agent pulls the terminal output and error logs. You can read these to see exactly why a test failed.
DIFFS
Code change validation
The AI compares the new code against your requirements. It identifies if the logic actually matches what you asked for.
HUMAN
Manual review needed
If the tests fail or the logic is messy, the AI stops. You decide whether to fix the code or reject the delivery.
The workflow
What your AI does when code is submitted.
The agent moves from writing code to proving it works.
Execute tests
The agent triggers your test runner to check the new code in your environment.
verify_deliveryCollect output
The agent captures the stdout and stderr from the test run to find errors.
verify_deliveryValidate logic
The agent inspects the diff to ensure the changes meet your specific instructions.
verify_deliveryReport findings
The agent presents the test results and logs so you can verify the work.
verify_delivery
Try it
Copy these to start.
Use these prompts to kick off a verification task.
Starting points
These are opening lines. Swap out the repo names or branch names to match your project.
Delivery Integrity Prover Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_47XZJ7Jwbw4wbyoAk0N57vFtYyOnkyNmEkgNiNe8/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Delivery Integrity Prover capabilities are ready to use.
{
"mcpServers": {
"delivery-integrity-prover-mcp": {
"url": "https://edge.vinkius.com/vk_preview_47XZJ7Jwbw4wbyoAk0N57vFtYyOnkyNmEkgNiNe8/mcp"
}
}
}Verify the changes in the current git branch and show me the test logs.
Run the test suite for the auth module in my repo and confirm it passes.
Check the diff for my last commit against the requirements in the README.
Run all tests for the user-service repo and provide proof of successful execution.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Delivery Integrity Prover once, then ask.
Click the link to connect the MCP to your client. Your credentials stay encrypted on our side and you can start verifying code immediately.
Connect Delivery Integrity Prover to your AIFAQ
How the verification behaves
- 01
Does the AI automatically merge my code?
No. The AI only verifies the code and provides the proof. You must manually merge the code in your git client.
- 02
Can the AI fix the code if the tests fail?
The AI can suggest fixes, but it won't rewrite your files without your permission.
- 03
What kind of test logs does it provide?
It provides the standard output and error logs generated by your test runner.
- 04
Does this MCP change my local files?
No. It reads your code and runs tests, but it does not modify your source files unless you tell it to.
- 05
Can it run tests on a remote server?
It runs tests through the environment your AI client has access to via the verify_delivery tool.
More questions about Delivery Integrity Prover? The Connector page answers them. See everything the Delivery Integrity Prover Connector can do
Connect Delivery Integrity Prover to Claude, Cursor, ChatGPT & more
