pm-verify

To verify acceptance criteria for project phases or work packages, you can use automated tests, manual code review, or AI assessment. The tool parses target IDs, loads context via MCP, and executes verification logic to confirm work meets defined standards.

The verification process does not alter entity states when recording results. It integrates with MCP tools to log verification outcomes without modifying the underlying project management entities, ensuring a non-destructive workflow.

The best way to run automated tests and code reviews for new features is to execute relevant test suites and parse results, then guide manual code reviews or leverage AI to assess code against complex criteria before completion.

Yes, you can use this to perform an AI assessment of code against specific criteria. The agent review feature leverages AI to evaluate code against complex project requirements during the verification process.

You need MCP tools to execute verification logic for acceptance criteria because the system relies on them to load context and record verification results without altering entity states.

Using a dry-run flag for verification limits the process to parsing arguments and loading context via MCP without executing the actual verification logic or recording the final results into the project management system.