• SUCCESS
  • FAILURE
  • HUMAN
  • RETRY

How to validate engine execution traces with your AI

Connect QA Arbiter to your agent. Once connected, your agent can inspect execution paths to prove code is working as intended.

Ask AI about this page

Short answer

How do I force my AI to provide a mathematical trace of code execution?

You give your agent access to execution logs and design specs. It uses diagnostic tools to map the actual runtime path against your expected logic. You get a verifiable proof of correctness.

Validation outcomes

Where your results land.

The AI processes the trace and categorizes the outcome based on the logic check.

SUCCESS

Mathematical proof

The AI confirms the execution path matches your design. You get a clean trace showing the logic held true.

FAILURE

Logic mismatch

The agent finds where the code diverged from the spec. It pinpoints the exact step where the execution failed the design test.

HUMAN

Manual review

When the trace is ambiguous or the logic is too complex, the AI flags it for you to inspect the raw logs yourself.

RETRY

Trace refinement

If the logs are incomplete, the AI asks for more granular execution data to finish the validation.

The workflow

What your AI does when a test fails.

The agent moves from detecting a failure to providing a formal proof of why it happened.

  1. Identify divergence

    The agent looks at the failed test and compares the actual execution path to your intended design.

    diagnose_test_failure

  2. Map the trace

    The AI builds a step-by-step mathematical representation of the engine's movement through the code.

    diagnose_test_failure

  3. Verify against specs

    The agent checks every state change in the trace against your formal requirements.

    diagnose_test_failure

  4. Report the proof

    The AI delivers a final report showing exactly where the execution trace violated the design.

    diagnose_test_failure

Try it

Copy these to start.

Use these prompts to kick off the validation process.

Starting points

Swap the bracketed text for your specific file paths or test names.

QA Arbiter Connector

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

Connector linkhttps://edge.vinkius.com/vk_preview_X69TXPHsdjAfNffcyFTnx1dWNUhv3RVHL0mSZM7p/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 — QA Arbiter capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "qa-arbiter-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_X69TXPHsdjAfNffcyFTnx1dWNUhv3RVHL0mSZM7p/mcp"
    }
  }
}
Copy into chat04

  • Run a trace validation on the engine logic in /src/core/engine.py.

  • Use the diagnostic tool to find why the last test in the suite failed.

  • Compare the execution trace of the current build against the design spec in /docs/spec.md.

  • Provide a mathematical proof that the state machine in the engine is behaving correctly.

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

Start here

Connect QA Arbiter once, then ask.

Just link your preferred client to the MCP. Your credentials stay encrypted, and you can start asking for traces immediately.

Connect QA Arbiter to your AI

FAQ

How the validation behaves

  • 01

    Can the AI fix the code automatically?

    No. The AI identifies the mismatch and provides the trace, but you must implement the actual code fix.

  • 02

    Does the AI change my design specifications?

    No. The AI treats your specs as the ground truth and only reports on how the code deviates from them.

  • 03

    What if the execution logs are missing data?

    The AI will tell you it cannot complete the trace and will ask you to provide more detailed logs.

  • 04

    Can I use this with any IDE?

    Yes, as long as your IDE is an MCP-compatible client like Cursor or VS Code.

  • 05

    Does the AI delete my test files?

    No. The AI only reads the traces and logs to perform its analysis.

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

Connect QA Arbiter to Claude, Cursor, ChatGPT & more