What happens if the design is insufficient for a task?

The skill stops immediately, informs the user of the gap, and waits for the design document to be updated before proceeding to prevent implementation errors or guesswork.

How does it handle interrupted sessions?

It reconstructs the current state by reading the tasks.md file within the feature directory, identifying which tasks are unblocked and not yet started or in progress.

What file structure does this skill require?

It requires a feature directory under docs/features/ containing three files: requirements.md, design.md, and tasks.md, typically generated by the spec-create skill.

Can it add new dependencies during implementation?

If a dependency is explicitly mentioned in the design document, it will install it; otherwise, it treats it as a design gap and asks the user for confirmation before proceeding.

Does it support Test-Driven Development (TDD)?

Yes, if the verification command is a recognized test runner (like Jest, Pytest, or Cargo test), the skill follows a strict red-green-refactor cycle by writing the test first.

Spec-Driven Implementer

Name: Spec-Driven Implementer
Author: jrollin

byjrollin

0•

Productivity & Workflow

Executes development tasks atomically from a technical specification to ensure high-fidelity feature implementation.

Spec-Driven Implementer is a specialized Claude Code skill designed to transform structured technical specifications into production-ready code. By reading requirements, design documents, and task lists from a feature directory, it executes implementation steps one-by-one with strict adherence to design decisions and automated verification. It bridges the gap between planning and coding by maintaining state, verifying design sufficiency, and enforcing test-driven development (TDD) cycles, making it ideal for developers who want to maintain perfect traceability between their technical specs and their codebase.

Key Features

01Multi-language support for automated verification commands and test runners

02Atomic task execution with persistent state management via tasks.md

03Automatic design sufficiency checks to prevent implementation guesswork

04Integrated Test-Driven Development (TDD) workflow for test-based verification

050 GitHub stars

06Context-aware implementation using requirements, design, and file inventory files

Use Cases

01Ensuring 100% test coverage and design compliance during the coding phase

02Implementing a new complex feature from a pre-defined technical specification

03Resuming a multi-day implementation session with automated state recovery

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jrollin/claudio spec-impl

For use in Claude.ai and ChatGPT

Download Skill