docs-manager-skill

Orchestrate single-document write, validate, and index workflows with rollback.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/fractary/claude-plugins --skill docs-manager-skill
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: docs-manager-skill
Source: https://github.com/fractary/claude-plugins/tree/main/plugins/docs/skills/docs-manager-skill
Command: npx skills add https://github.com/fractary/claude-plugins --skill docs-manager-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, bash, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill orchestrates a complete single-document workflow: write, validate, and index, ensuring one doc at a time with proper validation and indexing.

Core Features & Use Cases

  • Write a document with a type-aware context
  • Validate structure and content against type rules
  • Index updated docs for quick search and retrieval

Quick Start

Initiate a write → validate → index pipeline for a single document path.

Frequently Asked Questions about docs-manager-skill

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

FAQPage Schema
How do I automate a complete document workflow from write to index?▼

This Skill orchestrates a full single-document workflow: write with type context, validate structure and content, then index for retrieval. It handles the entire write→validate→index pipeline in one coordinated process, with rollback protection if indexing fails.

Can I validate documents against type-specific rules automatically?▼

Yes. The Skill loads type context from types/{doc_type}, then invokes the doc-validator to check structure and content against your type rules before indexing, ensuring only valid documents enter your search system.

What happens if validation or indexing fails during the workflow?▼

The Skill includes explicit error handling and automatic rollback on index failure, so invalid or failed documents don't corrupt your index. Failures are reported in structured results for downstream debugging.

How do I process one document at a time through a write-validate-index pipeline?▼

Provide a single document path and doc_type. The Skill loads type context, executes write→validate→index steps in sequence with optional skip controls, and outputs structured results for RAG systems or further processing.

Do I need bash and jq to run this Skill?▼

Yes. The Skill depends on bash for orchestration and jq for structured data processing. Both are required to execute the write→validate→index pipeline and handle JSON results.

What's the difference between running isolated steps versus this coordinated workflow?▼

This Skill synchronizes write, validate, and index as a single transaction with type-aware context and rollback safety. Running steps separately risks inconsistent state, missing validation, or orphaned index entries.