docs-writer

Create and maintain repository documentation including ADRs, runbooks, and READMEs.

3|1|Updated Apr 12, 2026
One-click install
npx skills add https://github.com/Cogni-AI-OU/cogni-ai-agent-skills --skill docs-writer-cogni-ai-ou
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: docs-writer
Source: https://github.com/Cogni-AI-OU/cogni-ai-agent-skills/tree/main/docs-writer
Command: npx skills add https://github.com/Cogni-AI-OU/cogni-ai-agent-skills --skill docs-writer-cogni-ai-ou

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

docs-writer eliminates inconsistent, outdated, or missing documentation by giving agents a clear standard for writing architecture-level rationale and operational guidance that stays current.

Core Features & Use Cases

  • Architectural Decision Records (ADRs): Capture context, decisions, alternatives, consequences, and status for significant architectural choices (often in docs/decisions/).
  • Runbooks: Define when humans use a process, the steps to execute, what “done” looks like, and how to recover from common failures (typically in docs/runbooks/).
  • README and code-level documentation: Keep repository documentation accurate and explain non-obvious code decisions, edge cases, and external constraints.
  • Subdirectory AGENTS.md: Add targeted local instructions that align with global invariants for agents operating within specific folders.

Quick Start

Use the docs-writer skill to create or update the project documentation requested in your task, including any ADRs and runbooks needed for the change you are implementing.

Frequently Asked Questions about docs-writer

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

FAQPage Schema
How do I write an Architecture Decision Record that captures context and consequences?▼

To write an Architecture Decision Record (ADR), document the context, decision, alternatives, consequences, and status fields to capture the rationale and impact of significant architectural choices in your repository.

What should a production runbook include for operational clarity?▼

A production runbook should define when humans use the process, the exact steps to execute, what "done" looks like, and how to recover from common failures, ensuring operational clarity for production run documentation.

How do I keep repository README and code-level documentation current?▼

Keep repository README and code-level documentation accurate by explaining non-obvious code decisions, edge cases, and external constraints, applying documentation quality rules focused on the "why" behind the code.

Can I add targeted local instructions for agents operating within specific folders?▼

Yes, you can add subdirectory AGENTS.md files that provide targeted local instructions for agents operating within specific folders, ensuring those instructions align with global invariants for the repository.

When do I need to create an ADR for architecture changes?▼

You need to create an ADR for architecture changes when a significant architectural choice is made, capturing the context, decision, alternatives, consequences, and status to maintain reasoning and operational clarity.

What is the best way to document recovery steps for common production failures?▼

The best way to document recovery steps for common production failures is to create a runbook that defines the execution steps, what "done" looks like, and specific recovery procedures to maintain operational clarity.