documentation-authoring

v2026.09.24

Create structured docs from scratch — PRDs, technical specs, design docs, decision records, knowledge bases. Use when drafting documentation, writing proposals, defining requirements, or planning features.

GitHub
安装命令
npx skhub add practicalswan/documentation-authoring
Markdown
SKILL.md

Documentation Authoring Master

Expert guidance for creating structured, high-quality documentation across all types of technical and business documents.

  • Leverage native parallel subagent dispatch and 200k+ context windows where available.

Activation Conditions

Use symptom -> action triggers: when one matches, apply this skill and verify with the protocol below.

Trigger Conditions:

  • User mentions writing documentation: "write a doc", "draft a proposal", "create a spec", "write up"
  • User mentions specific doc types: "PRD", "design doc", "decision doc", "RFC"
  • User asks to "create an implementation plan", "document requirements", "plan a feature"
  • Creating technical specifications or business requirements
  • Starting a new product or feature development cycle
  • Translating vague ideas into concrete technical specifications
  • Stakeholders need unified "source of truth" for project scope

Part 2: Context Gathering

Initial Questions

Start by asking for meta-context about the document:

  1. What type of document is this?

    • Technical spec, decision doc, proposal, RFC, PRD, knowledge base
  2. Who's the primary audience?

    • Developers, executives, stakeholders, end-users? Understanding affects tone and depth
  3. What's the desired impact when someone reads this?

    • Make a decision, implement a feature, understand a concept?
  4. Is there a template or specific format to follow?

    • Company templates, industry standards, regulatory requirements
  5. Any other constraints or context to know?

    • Deadlines, sensitive information, existing related documents

Inform them they can answer in shorthand or dump information however works best for them.

Template Handling

If user provides a template:

  • Analyze structure and requirements
  • Adapt co-authoring workflow to template format
  • Ensure all required sections are covered

If user mentions editing an existing document:

  • Fetch the existing document
  • Understand current state and gaps
  • Plan revisions strategically

Part 3: Refinement & Structure

Collaborative Building

Process:

  1. Brainstorm each section together - let ideas flow without judgment
  2. Organize and refine - structure ideas into coherent sections
  3. Edit for clarity - improve readability and flow
  4. Add professional polish - formatting, consistency, tone

Guiding Principles:

  • Active voice: Use direct, clear language
  • Show, don't just tell: Use examples and scenarios
  • Progressive disclosure: Start with overview, then dive deeper
  • Visual aids: Include diagrams, tables, and examples where helpful

Section-by-Section Approach

Work through document methodically:

## Recommended Section Structure

### 1. Executive Summary (for decision-makers)
- What is this about?
- Why does it matter?
- What are we recommending/deciding?

### 2. Background & Context (for implementers)
- What led us here?
- What problem are we solving?
- What constraints exist?

### 3. Requirements/Objectives
- What must we achieve?
- What are success criteria?
- What are non-goals?

### 4. Proposed Solution/Design
- What are we proposing?
- How does it work?
- What are alternatives considered?

### 5. Implementation Plan
- How do we build this?
- What are the steps?
- Who needs to do what?

### 6. Risks & Considerations
- What could go wrong?
- How do we mitigate?
- What decisions are still needed?

Part 4: Reader Testing

The Fresh Eye Test

Before finalizing, put yourself in the reader's shoes:

Test Questions:

  1. Can I understand the goal without knowing context?
  2. Are technical terms explained or linked?
  3. Is there a logical flow from problem to solution?
  4. Would a skeptical reader be convinced?
  5. Is action clear - what should I do next?

Blind Spot Detection

Common issues to catch:

  • Context assumptions: "We already discussed this" but wasn't documented
  • Missing alternatives: Only one option presented (shows lack of thoroughness)
  • Unanswered questions: Reader left with "what about X?"
  • Unclear responsibilities: Who needs to do what is vague
  • Missing examples: Abstract concepts without concrete illustration

Part 5: Product Requirements Document (PRD)

PRD Structure

When users specifically request PRDs or feature planning, use this structure:

# [Feature/Product Name] - PRD

## Executive Summary
**Goal**: [What problem are we solving?]
**Impact**: [Why does this matter now?]
**Success Metrics**: [How will we know it worked?]

## Background
**Current State**: [What's the situation today?]
**Problem Statement**: [What pain points exist?]
**Constraints**: [Budget, timeline, tech stack limitations?]

## Requirements

### Functional Requirements
- User stories with acceptance criteria
- Core features and capabilities
- Integration requirements

### Non-Functional Requirements
- Performance requirements
- Security requirements
- Compliance and regulatory needs

### User Stories

As a [persona], I want to [action], So that [benefit].

**Acceptance Criteria**:
- [ ] [Specific, measurable criterion]
- [ ] [Another criterion]

Proposed Solution

Architecture Overview

[High-level system architecture or approach]

Technical Specifications

[API contracts, data models, interfaces]

UI/UX Requirements

[Wireframes or flow descriptions if applicable]

Implementation Plan

Phases

PhaseTasksOwnersTimeline
Phase 1
Phase 2

Dependencies

  • External APIs or services
  • Other teams or systems
  • Third-party libraries

Risk Analysis

RiskImpactProbabilityMitigation
[Risk]High/Med/LowHigh/Med/Low[Mitigation]

Alternatives Considered

OptionProsConsWhy Not Chosen
Alt 1

Success Criteria

Quantitative

  • [Measurable metric: e.g., "reduce load time by 50%"]
  • [Another metric]

Qualitative

  • [User feedback threshold]
  • [Stakeholder alignment]

Open Questions

  • [Decision still needed]
  • [Information to gather]

### PRD Creation Workflow

**Phase 1: Discovery (The Interview)**
Before writing a single line, you **MUST** interrogate user to fill knowledge gaps. Do not assume context.

**Ask about:**
- **The Core Problem**: Why are we building this now?
- **Success Metrics**: How do we know it worked?
- **Constraints**: Budget, tech stack, or deadline?
- **Stakeholders**: Who needs to approve? Who will use?

**Phase 2: Analysis & Scoping**
Synthesize user input. Identify dependencies and hidden complexities.
- **Map out User Flow**
- **Define Non-Goals** to protect timeline

**Phase 3: Technical Drafting**
Generate document using strict structure above.

---

## Part 6: Common Document Types

### Implementation Plans

**Purpose**: Guide building process with clear phases, responsibilities, and timeline.

**Structure:**
- **Overview**: What are we building and why?
- **Phases**: Break into logical chunks with dependencies
- **Tasks**: Trackable, specific implementation items
- **Timeline**: Realistic dates with buffers
- **Dependencies**: What must happen before what?

### Design Docs

**Purpose**: Document technical decisions and architecture.

**Structure:**
- **Problem Statement**: What problem are we solving?
- **Alternatives**: What did we consider?
- **Decision**: What did we choose and why?
- **Implications**: What does this mean for the system?
- **Risks**: What could go wrong?

### Decision Records

**Purpose**: Capture important decisions for future reference.

**Template:**
```markdown
### Decision - [DATE]
**Decision**: [What was decided]
**Context**: [Situation and driving data]
**Options**: [Alternatives with pros/cons]
**Rationale**: [Why selected option is superior]
**Impact**: [Anticipated consequences]
**Review**: [Reassessment conditions/trigger]

Knowledge Base Articles

Purpose: Reusable reference material, not project-specific docs.

Structure:

  • Quick Reference: TL;DR summary at top
  • Problem: What question does this answer?
  • Solution: How do you solve it?
  • Examples: Concrete, runnable examples
  • Common Pitfalls: What mistakes do people make?
  • Related Topics: Links to related info

Part 7: Best Practices

For All Documentation

✅ DO:

  • Use active voice and clear language
  • Structure information progressively (simple to complex)
  • Include examples and concrete scenarios
  • Define terms before using them
  • Add diagrams for complex systems
  • Maintain consistent formatting and style

❌ DON'T:

  • Write without clear audience in mind
  • Mix jargon without explanation
  • Skip alternatives or trade-offs analysis
  • Assume readers have context they don't
  • Create long paragraphs without breaks

For Technical Docs

  • Include code snippets that actually run
  • Link to external references for deeper dives
  • Use standard terminology when possible
  • Version specific code/commands (e.g., "for node v16+")

For Business/Stakeholder Docs

  • Start with executive summary
  • Use business impact metrics
  • Hide unnecessary technical detail
  • Include clear next steps or approvals needed
  • Highlight risks and mitigations prominently

Part 8: Action Documentation Format

Use this format for tracking implementation work and decisions:

[TYPE] - [ACTION] - [TIMESTAMP]

Objective: [Goal being accomplished]

Context: [Current state, requirements, reference to prior steps]

Decision: [Approach chosen and rationale]

Execution: [Steps taken with parameters and commands]

Output: [Complete results, logs, metrics]

Validation: [Success verification and results]

Next: [Continuation plan to next action]


Part 9: Summary Formats

Streamlined Action Log (for changelogs)

[TYPE][TIMESTAMP] Goal: [X] → Action: [Y] → Result: [Z] → Next: [W]

Quick Summary (for updates)

What: [Brief description] Why: [Context/rationale] How: [Approach taken] Status: [Current state] Next: [Upcoming step]


Documentation Stack Reference

Inherit the shared stack from documentation-patterns: source-of-truth discovery, audience framing, structure selection, verification, and freshness checks. Keep this skill focused on drafting and refinement instead of restating the full stack.

<!-- MCP:START --> <!-- PORTABILITY:START -->

Cross-Client Portability

This skill is written to stay usable across GitHub Copilot, Claude Code, and Codex.

  • GitHub Copilot: keep the folder in a Copilot-visible skill path or wrap the workflow in project instructions when folder discovery is unavailable.
  • Claude Code: keep the folder in a local skills directory or a compatible plugin source.
  • Codex: install or sync the folder into $CODEX_HOME/skills/documentation-authoring and restart Codex after major changes.
<!-- PORTABILITY:END -->

MCP Availability And Fallback

Preferred MCP Server: None required

  • Fallback prompt: "Use the Documentation Authoring Master skill without MCP. Rely on its local instructions, bundled resources, standard shell or editor tools, and direct verification. Show the evidence used before concluding."
  • Do not claim an MCP operation was used when the active host does not expose it.
  • Treat local files, tests, rendered outputs, logs, or screenshots as the fallback evidence path.
<!-- MCP:END -->

Anti-Patterns

  • Writing for the author instead of the reader: It bakes in unstated context and leaves the actual audience unsure what to do next.
  • Skipping concrete examples or commands: Abstract guidance is easy to approve and hard to apply correctly.
  • Letting links, screenshots, or versions drift: Polished formatting does not help if the instructions are no longer true.

Verification Protocol

Before claiming "skill applied successfully":

  1. Pass/fail: The Documentation Authoring output identifies audience, purpose, source of truth, and freshness requirements.
  2. Pass/fail: Shared documentation-stack guidance is referenced instead of duplicating another documentation skill.
  3. Pass/fail: Claims, links, commands, examples, and screenshots are verified or explicitly marked unverified.
  4. Pressure-test scenario: Apply the skill to a doc request with a stale command, missing owner, and conflicting audience.
  5. Success metric: Zero undocumented assumptions; every reader-facing claim is sourced or scoped.

Documentation Quality Checklist

Completeness

  • All required sections filled
  • Context and background provided
  • Alternatives considered where applicable
  • Examples and diagrams included where helpful

Clarity

  • Language is clear and direct
  • Technical terms defined or linked
  • Flowlogical and easy to follow
  • Active voice used consistently

Accuracy

  • Technical details are correct
  • Links work and are up-to-date
  • Code examples actually run
  • No contradictory information

Accessibility

  • Multiple levels of detail for different readers
  • Executive summary for decision-makers
  • Deep-dive sections for implementers
  • Visual aids for complex concepts

References & Resources

Documentation

  • Document Templates — Templates for PRD, RFC, ADR, Tech Spec, Design Doc, Runbook, Postmortem, KB Article
  • Writing Style Guide — Technical writing best practices, formatting conventions, and readability

Scripts

Examples

  • PRD Example — Complete PRD for Recipe Search Enhancement in Kitchen Odyssey

Cross-Skill Workflow

  • Start in this skill when you need to draft or reshape the document itself.
  • Pull in documentation-patterns when the structure or template is the main decision.
  • Finish with documentation-quality when the draft needs an explicit review against quality bars.

Agent Prompt Template

Use the documentation-authoring skill to draft a [document type] for [audience].
Goal: [decision, rollout, implementation, or explanation target].
Required sections: [list].
Constraints: [scope, timeline, compliance, or tooling notes].
Include open questions, trade-offs, and next steps at the end.

Related Skills

  • documentation-patterns: Use it when the workflow also needs reusable documentation structures and templates.
  • documentation-quality: Use it when the workflow also needs documentation review standards and quality gates.
  • documentation-verification: Use it when the workflow also needs final documentation validation before publishing.
  • notion-docs: Use it when the workflow also needs Notion page and database publishing workflows.
发现
标签

此技能尚未发布标签。

版本
最新版本元数据

版本

v2026.09.24

发布时间

2026年9月24日

分类

未分类

许可证

MIT

源路径

documentation-authoring

默认分支

main

最新提交

ff6d12f

Tree SHA

e96fd60