claude-task-completion-verifier

claude-task-completion-verifier MCP Connector for Claude

A+

A deterministic verification engine that validates task fulfillment by checking files, content, and command execution.

3 tools Official Updated Oct 1, 2026 Official Vinkius Partner

This MCP server provides a deterministic verification engine to ensure AI agents have truly completed their tasks. Instead of relying on intuition, it uses objective evidence by checking for file existence, matching regex patterns, and executing shell commands like tests or builds. Use verify_task_completion to perform a full audit, check_file_integrity for quick file and pattern checks, or execute_environment_check to run technical health commands. It bridges the gap between task description and verifiable reality.

verificationdeterministictestingautomationaudit

3 tools expose this connector's capabilities to your AI agent.

verify_task_completion

Performs a full audit of a task by evaluating all provided criteria

execute_environment_check

Runs shell commands to verify the technical health of the workspace

check_file_integrity

Verifies that specific files exist and contain expected content patterns

See how to talk to your AI agent using claude-task-completion-verifier.

Verify if the file 'src/engine/index.ts' exists and contains the word 'verify_task_completion'.

The file exists and the pattern matches successfully.

Run the tests in the workspace to see if they pass.

Tests passed successfully with exit code 0.

Perform a full audit of the task: 'Implement the verification engine'.

Task is 100% complete based on all provided criteria.

The engine uses `verify_task_completion` to evaluate specific criteria such as file existence, content matching, or command exit codes to provide a mathematical completion percentage.

Related Connectors