Task Completion Enforcer Prover Connector for AI agents.

It forces the agent to identify every requirement as a checklist. If it finds a gap or a placeholder, it is required to fix it immediately before it can move to the next step.

Yes. It is specifically designed for multi-step coding projects where missing a single detail like a test case or a specific documentation update causes a headache.

It works with any MCP-compatible client, including Cursor, VS Code, and Claude, allowing you to use it directly within your existing development environment.

LLMs often prioritize finishing a generation over finishing a task. This Connector forces the agent to provide verifiable evidence for every requirement before it is allowed to say it is done.

The agent must link every finished requirement to a specific artifact, such as a file path, a line number, or a specific test result, to prove the work was actually completed.

Yes. It treats documentation as a mandatory requirement. The agent cannot declare the task finished until it provides evidence that the docs have been updated as requested.

Autoregressive generation allocates decreasing attention to earlier tokens as output grows. A 10-step task at token 200 competes with 2,000 tokens of generated output for attention. The model literally loses track of requirement #7 while implementing requirement #3. The fix: force a re-read of ALL requirements before declaring completion.

Not 'I implemented the function.' Evidence means: 'Requirement 1: POST /users endpoint at src/routes/users.ts lines 15-42, validates email/name/role via Zod schema, returns 201 with user object.' File path, line number, specific behavior. If you cannot point to the exact artifact, it is not done.

The LLM MUST close them immediately. not later, not in a follow-up. Do the remaining work NOW. Then call this capability AGAIN to verify the gaps are actually closed. The loop continues until EVERY requirement has concrete evidence. 'I will do it later' is rejected. 'I just did it, here is the evidence' is accepted.