documentation-librarian

Maintain repository documentation as a versioned knowledge graph with drift detection.

Updated Aug 23, 2026

Search Tags:#documentation #governance #drift-detection #knowledge-graph #versioned-docs #subtree-repair

Ready to run

npx skills add https://github.com/claushaas/claus-haas-ai-stuff --skill documentation-librarian

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: documentation-librarian
Source: https://github.com/claushaas/claus-haas-ai-stuff/tree/main/skills/documentation-librarian
Command: npx skills add https://github.com/claushaas/claus-haas-ai-stuff --skill documentation-librarian

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The Documentation Librarian preserves correctness, consistency, and continuity across a repository's documentation by treating docs as a connected, versioned knowledge graph rather than a collection of isolated files.

Core Features & Use Cases

Graph-based governance: maintains a coherent documentation graph and detects drift across README, SPEC, Reference, ADRs, and roadmaps.
Incremental repairs: updates the necessary subtree of documents to preserve global consistency without rewriting unrelated content.
Mode-based operation: supports Bootstrap, Relationship Repair (Mode 2B), and Controlled Refactor workflows; generates change reports and flags risks.
Operational guidance: outputs minimal edits with explicit tags ([ASSUMPTION], [PENDING], [RISK], [SOURCE]) to preserve traceability.
Use cases: stabilize cross-document invariants, align boundaries between docs, and verify document graph integrity.

Quick Start

Inventory the repo's docs graph using the provided graph index or map.
Identify drift between documents (README, SPEC, Reference, ADRs) and determine affected subtree.
Run a Mode 2B subtree repair to align the subtree, then generate a concise change report with explicit tags.

Frequently Asked Questions about documentation-librarian

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

FAQPage Schema

How do I prevent documentation drift across multiple repository files?▼

Preventing documentation drift requires treating repository docs as a connected, versioned knowledge graph rather than isolated files. This approach detects inconsistencies across README, SPEC, Reference, and ADRs, ensuring cross-document invariants remain stable during changes.

What is graph-based documentation governance?▼

Graph-based documentation governance maintains a coherent knowledge graph to detect drift and align boundaries across README, SPEC, Reference, ADRs, and roadmaps. It preserves correctness and continuity by applying minimal subtree edits instead of rewriting unrelated content.

How do I repair inconsistent documentation subtrees without rewriting everything?▼

To repair inconsistent documentation subtrees without rewriting everything, run an incremental subtree repair workflow. This updates only the necessary affected documents to preserve global consistency, generating a concise change report with explicit risk and assumption tags.

When do I need to bootstrap a documentation knowledge graph?▼

You need to bootstrap a documentation knowledge graph when stabilizing cross-document invariants for the first time or aligning boundaries between README, SPEC, and Reference files. This establishes a coherent baseline to detect drift and verify document graph integrity.

Can I use a knowledge graph approach for controlled documentation refactors?▼

Yes, you can use a knowledge graph approach for controlled documentation refactors. It supports controlled refactor workflows that verify document graph integrity, flag risks explicitly, and apply minimal edits to maintain consistency across ADRs and roadmaps.

How does documentation graph governance handle assumptions and risks?▼

Documentation graph governance handles assumptions and risks by outputting minimal edits with explicit tags like [ASSUMPTION], [RISK], [PENDING], and [SOURCE]. This traceability ensures flagged risks and assumptions are visible in the concise change report after subtree repairs.