How does this skill save tokens during AI development?

It uses a progressive disclosure pattern where only metadata is loaded initially; Claude then uses search commands to read only the specific files necessary for the current task.

What is a 'living document' in this context?

Living documents are frequently updated files stored within your repository that capture requirements, architecture decisions, and operational guides as the software evolves.

Where should I store my project documents?

Documents should be organized within the .specweave/docs/internal/ directory, utilizing subfolders for specs, architecture, ADRs, and operations.

Do I need a RAG system for this skill to work?

No, this skill leverages Claude's native file-reading capabilities and directory navigation, which provides higher accuracy for local codebase documentation than traditional RAG.

Can I use this with existing documentation?

Yes, by moving your existing markdown documentation into the Specweave directory structure, the Living Docs Navigator can immediately index and utilize it.

Living Docs Navigator

Name: Living Docs Navigator
Author: anton-abyzov

byanton-abyzov

•

学习与文档

Navigates and retrieves project context from living documentation including specifications, architecture decisions, and runbooks.

The Living Docs Navigator is a specialized skill for Claude Code that enables autonomous exploration of a project's internal knowledge base within the Specweave framework. By organizing documentation into structured directories for specs, ADRs, and system architecture, it allows the AI to gain deep domain context, understand historical design trade-offs, and align implementation with established governance. It utilizes a progressive disclosure pattern to minimize token usage, ensuring Claude only reads the most relevant documentation required for the task at hand.

主要功能

01Deep integration with Specweave autonomous development workflows

02Navigation of feature specifications, user stories, and acceptance criteria

03Contextual search using grep for rapid discovery of internal project knowledge

04Progressive disclosure pattern to maximize context while minimizing token cost

05Structured access to Architecture Decision Records (ADRs) for design rationale

0623 GitHub stars

使用场景

01Verifying historical design decisions to avoid contradicting project patterns

02Onboarding an AI agent to a complex codebase via existing architecture docs

03Gathering requirement context before implementing new features or bug fixes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add anton-abyzov/specweave living-docs-navigator

For use in Claude.ai and ChatGPT

Download Skill