wf-do

Execute document-defined development tasks from spec.md and tasks.md across repositories.

2|Updated Apr 23, 2026
One-click install
npx skills add https://github.com/liqunx/wikiflow --skill wf-do
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wf-do
Source: https://github.com/liqunx/wikiflow/tree/main/skills/wf-do
Command: npx skills add https://github.com/liqunx/wikiflow --skill wf-do

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns a document-driven development plan into real code changes by executing tasks sequentially based on spec.md and tasks.md, reducing the gap between plans and implementation.

Core Features & Use Cases

  • Plan-aware task execution: Reads spec.md to understand scope and tasks.md to determine what to do next.
  • Cross-repository execution with guardrails: Executes in the documentation repository while modifying only the code repository via configured paths.
  • Progress visibility and failure control: Reports completion progress and stops on failures to ask whether to continue, keeping the workflow aligned with the plan.
  • Use Case: When adding a new feature (e.g., “registration”), run wf-do to implement backend and frontend tasks one by one according to the active plan.

Quick Start

Run wf-do in your documentation repository to execute the currently active plan from spec.md and tasks.md.

Frequently Asked Questions about wf-do

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute development tasks defined in markdown documentation?

To execute development tasks defined in markdown documentation, you can use a specification-driven workflow tool to read spec.md and tasks.md, iterating through plan items to apply code changes sequentially.

How does specification-driven development work with separate documentation and code repositories?

Specification-driven development with separate repositories works by applying strict read and write boundaries. The tool reads the plan from the documentation repository while modifying only the linked code repository via configured paths.

What is the best way to track task completion progress across documentation and code repositories?

The best way to track task completion progress is to use a plan execution workflow that updates task status in the documentation repository while implementing the actual code changes in the linked code repository.

Do I need a config.json file to resolve language and project settings for task execution?

Yes, you need a config.json file to resolve language and project settings. The workflow requires this configuration to enforce strict read and write boundaries between the documentation and code repositories.

How do I handle task execution failures when implementing a plan sequentially?

To handle task execution failures, the workflow stops on failures and asks whether to continue. This controlled failure handling keeps your development progress aligned with the active plan and prevents uncontrolled errors.