Skill: Spec Implementer – Testmo

Spec Implementer turns your existing Testmo test cases into working context for an AI coding assistant. Before writing new code for a feature or change, it pulls the test cases that describe how the relevant part of the system is supposed to behave, and feeds them to the assistant — so the code it generates respects the rules your tests already encode.

This is a closed beta. The Testmo MCP is available to selected teams while we refine it toward general availability. If you'd like access, fill out the form here: Testmo MCP Beta

What it does

Your test cases capture business rules, edge cases, and expected behaviour that often aren't written down anywhere else — validation limits, status transitions, integration quirks. When an AI assistant writes code without that knowledge, it produces plausible-looking code that quietly violates those rules. Spec Implementer closes that gap:

You describe the change you want to make.
The skill finds the test cases relevant to that area of the system.
It distils them into the concrete requirements and constraints the code must satisfy.
The assistant implements the change with those requirements as guardrails.

When to use it

Implementing a new feature in an area that already has test coverage.
Modifying existing behaviour where you want the change to stay within established rules.
Working in an unfamiliar part of the codebase and wanting the tests to teach the assistant the domain rules first.

How to run it

Once the Testmo skills are installed (see the hub article), describe your task and let the skill gather context. For example:

Implement support for annual billing on the subscription page.
Use our Testmo cases for subscriptions and billing as the spec.

The skill uses Semantic Search to locate the relevant cases, summarises the behaviour they define, and hands that to the coding assistant as context for the implementation. A read-only connection is sufficient — Spec Implementer reads your cases, it doesn't change them.

What you get back

Generated or modified code that reflects the requirements captured in your existing tests, along with a summary of which cases informed the implementation — so you can see the reasoning and confirm nothing important was missed.