
Tech Writer Researcher
- 27 installs
- 7 repo stars
- Updated May 20, 2026
- daemon-blockint-tech/agentic-enteprises-skill
Write and research technical documentation: information architecture, style guides, API reference, user research, and content strategy.
About
Guides technical writing and research covering information architecture, style guides, API documentation, user research, content strategy, and documentation operations. Used when writing developer docs, API references, or building a documentation style guide.
- Build API reference with request/response examples and authentication guides
- Define information architecture, content models, and style guides
Tech Writer Researcher by the numbers
- 27 all-time installs (skills.sh)
- Ranked #951 of 1,879 Documentation skills by installs in the Skillselion catalog
- Data as of Jul 29, 2026 (Skillselion catalog sync)
npx skills add https://github.com/daemon-blockint-tech/agentic-enteprises-skill --skill tech-writer-researcherAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 27 |
|---|---|
| repo stars | ★ 7 |
| Last updated | May 20, 2026 |
| Repository | daemon-blockint-tech/agentic-enteprises-skill ↗ |
What it does
Write and research technical documentation: information architecture, style guides, API reference, user research, and content strategy.
Files
Technical Writer / Researcher
Overview
Write and research technical documentation. This skill covers information architecture, style guides, API documentation, user research, content strategy, and documentation operations.
Features
- Information architecture: content modeling, navigation design, taxonomy creation
- Style guides: voice and tone, terminology, code examples, accessibility standards
- API documentation: endpoint reference, request/response examples, authentication guides
- User research: persona development, usability testing, content gap analysis
- Content strategy: editorial calendars, content audits, migration planning
- Documentation operations: version control, review workflows, localization, metrics
Usage
1. Identify the user's technical writing need (architecture, style, API docs, research, or strategy) 2. Follow the corresponding workflow below 3. Produce structured outputs: content models, style guides, API reference docs, research reports, or content strategies
Examples
- User: "Write API documentation"
Agent: Runs API Documentation workflow, creates endpoint reference with request/response examples, authentication guide, and error codes
- User: "Create a style guide"
Agent: Runs Style Guide workflow, defines voice and tone, establishes terminology, produces code example standards
- User: "Audit our documentation"
Agent: Runs Content Strategy workflow, analyzes content gaps, identifies outdated pages, produces improvement roadmap
When to Use
- Authoring API references, tutorials, runbooks, READMEs, and release notes
- Planning documentation information architecture and content lifecycle
- Conducting structured research, literature reviews, and evidence synthesis
- Editing for clarity, consistency, accessibility, and style-guide compliance
When NOT to Use
- Prompt/LLM agent design, eval harnesses, or guardrails → use
prompt-engineer - Business requirements, BRDs, or process modeling for delivery projects → use
business-analyst - Warehouse SQL optimization or dimensional modeling → use
data-warehouse-engineer - Revenue accounting, close calendars, or ASC 606 judgments → use
senior-revenue-accountant - Customer ticket repro, escalation, support KB fixes → use
support-engineer - Product support how-tos, macros, ticket triage → use
product-support-specialist - Strategy consulting, executive recommendations, operating model → use
business-consultant - Business model and competitive monetization research → use
business-model-researcher - All-hands, crisis statements, launch messaging → use
communication-lead - Developer learning paths, workshops, certification programs → use
developer-education-lead
Core Workflows
1. Technical Documentation Workflow
Phase checklist:
1. Audience analysis
- Who reads this? (role, skill level, context)
- What do they need to do after reading?
- What do they already know vs. need to learn?
2. Information gathering
- Interview SMEs (subject matter experts)
- Review existing docs, code, and specs
- Test the product/feature yourself
3. Structure & outline
- Choose document type (see table below)
- Create heading hierarchy
- Identify prerequisites and next steps
4. Draft
- Write for clarity first, polish later
- Include code examples, screenshots, diagrams
- Use consistent terminology
5. Review
- Technical accuracy review (SME)
- Editorial review (style, grammar, clarity)
- User testing (if possible)
6. Publish & maintain
- Version control and changelog
- Feedback mechanism
- Scheduled review cadence
Document type selection:
| Type | Purpose | Audience | Length |
|---|---|---|---|
| README | Quick start, install, overview | Developers | 1-2 pages |
| API reference | Endpoint details, parameters | Developers | Per endpoint |
| Tutorial | Step-by-step learning | New users | Medium |
| How-to guide | Specific task completion | Users with context | Short |
| Explanation | Conceptual understanding | All levels | Medium |
| Runbook | Incident response, operations | On-call engineers | Short |
| Release notes | What changed, why | Users, stakeholders | Short |
| FAQ | Common questions | Support reduction | Varies |
2. Research & Synthesis Workflow
Structured research process:
1. Define the question
- Convert vague request into specific research question
- Identify scope (time period, geography, sources)
- Define success criteria (decision support, background, deep dive)
2. Source & collect
- Primary: interviews, surveys, observations
- Secondary: papers, reports, databases, news
- Tertiary: summaries, reviews, meta-analyses
3. Evaluate sources
- Currency, relevance, authority, accuracy, purpose (CRAAP test)
- Bias detection (funding, affiliation, methodology)
4. Synthesize
- Group findings by theme
- Identify agreements, contradictions, gaps
- Extract evidence-backed conclusions
5. Communicate
- Match format to audience (executive summary, full report, memo)
- Cite sources properly
- Include confidence levels and limitations
3. Content Strategy & Planning
Content inventory template:
| Document | Audience | Type | Owner | Last Updated | Review Cycle | Status |
|---|---|---|---|---|---|---|
| API Guide | Developers | Reference | @tech-writer | 2024-01 | Quarterly | Current |
4. Editing & Review
Self-editing checklist:
- [ ] One idea per paragraph
- [ ] Active voice (where appropriate)
- [ ] Defined acronyms on first use
- [ ] Consistent terminology
- [ ] Scannable headings and lists
- [ ] Working links and code examples
- [ ] Accessibility (alt text, color independence)
Content Strategy
Audience Analysis
Persona Template
## [Persona Name]
**Role:** Senior Developer at mid-size SaaS company
**Goals:** Evaluate if our API fits their integration needs
**Pain points:** Poor examples, unclear error messages, missing edge case docs
**Technical level:** Expert in their stack, beginner with our tech
**Context:** Evaluating during sprint planning, limited time
**Preferred format:** Copy-paste code examples, decision trees
**Quote:** "I just need to see if this handles webhooks properly."Audience Matrix
| Audience | Knowledge | Need | Format | Channel |
|---|---|---|---|---|
| Evaluators | Low | Quick assessment | README, quick start | GitHub, landing page |
| New users | Medium | Get running | Tutorials, guides | Docs site |
| Power users | High | Advanced patterns | Reference, blogs | Docs, GitHub |
| Contributors | Expert | Architecture, internals | Explanations, RFCs | Wiki, repo |
Content Planning
Content Calendar Template
| Week | Content | Type | Owner | Status | Channel |
|---|---|---|---|---|---|
| W1 | New feature guide | How-to | @writer | Draft | Docs |
| W2 | API changelog | Reference | @eng | Ready | GitHub |
| W3 | Architecture deep dive | Explanation | @architect | Idea | Blog |
Content Prioritization
Score each piece by:
- Reach: How many users affected?
- Impact: How critical is this information?
- Effort: Time to produce?
- Urgency: Is there a deadline or blocker?
Documentation Lifecycle
States
| State | Criteria | Action |
|---|---|---|
| Draft | Being written | Internal review |
| Review | SME checking accuracy | Incorporate feedback |
| Published | Live and discoverable | Monitor metrics |
| Current | Up-to-date, accurate | Schedule next review |
| Stale | Minor drift | Update in next cycle |
| Outdated | Significant drift | Priority update |
| Deprecated | Feature removed | Archive or redirect |
Review Cadence
| Doc Type | Review Frequency | Trigger |
|---|---|---|
| API reference | Per release | API version change |
| Runbooks | Quarterly | Incident post-mortem |
| Tutorials | Bi-annually | Product UI change |
| Architecture docs | Annually | Major redesign |
| README | Per release | Dependency update |
Content Metrics
| Metric | How to Measure | Target |
|---|---|---|
| Page views | Analytics | Trend up |
| Time on page | Analytics | 2-5 min (context-dependent) |
| Bounce rate | Analytics | <40% for landing pages |
| Search queries | Search logs | Match user vocabulary |
| Support tickets | Support system | Decline for documented issues |
| NPS/CSAT | Surveys | >7/10 |
| Contribution rate | GitHub metrics | Community PRs accepted |
Information Scent
Help users know they're in the right place:
- Page title matches search query
- First paragraph confirms relevance
- Headings preview what's covered
- Cross-links to related content
Content Reuse Strategy
Single Sourcing
Write once, publish everywhere:
- Source in Markdown/DITA
- Generate docs site, PDF, help center
- Include in product (in-app help)
Snippet Libraries
Maintain reusable content:
- Warnings and disclaimers
- Common troubleshooting steps
- Boilerplate definitions
- Code examples by language
Governance
RACI for Documentation
| Activity | Writer | SME | Editor | PM |
|---|---|---|---|---|
| Write first draft | R | C | I | I |
| Technical accuracy | C | R/A | I | I |
| Style review | R | I | R/A | I |
| Publish | R | I | C | A |
| Maintain | R | C | I | I |
Style Guide Governance
- Versioned style guide in repo
- Linting with Vale or similar
- Annual review for relevance
- Contribution process for additions
Multichannel Publishing
| Channel | Format | Update Frequency |
|---|---|---|
| Docs site | HTML/Markdown | Continuous |
| GitHub README | Markdown | Per release |
| In-app help | Embedded snippets | Per feature |
| Generated | Quarterly | |
| Video tutorials | Screencast | Per major feature |
| Blog posts | Markdown | Weekly/Bi-weekly |
| API explorer | Interactive | Real-time |
| Slack/Discord | Short snippets | As needed |
Research Methods
Research Design Frameworks
PICO (Evidence-Based Practice)
Used for clinical/technical literature reviews:
- Population: Who or what is being studied?
- Intervention: What action or exposure?
- Comparison: What is the alternative?
- Outcome: What is measured?
Example: "In distributed systems (P), does chaos engineering (I) vs traditional testing (C) reduce production incidents (O)?"
SPIDER (Qualitative Research)
- Sample: Who was studied?
- Phenomenon: What was explored?
- Design: How was it studied?
- Evaluation: What were the outcomes?
- Research type: Qualitative/quantitative/mixed
Source Evaluation (CRAAP Test)
| Criterion | Questions |
|---|---|
| Currency | When was it published? Is it still relevant? |
| Relevance | Does it answer your question? Is the level appropriate? |
| Authority | Who wrote it? What are their credentials? |
| Accuracy | Is it supported by evidence? Has it been reviewed? |
| Purpose | Why was it written? Is there bias? |
Research Types & When to Use
| Type | When | Output |
|---|---|---|
| Literature review | Understanding a field | Synthesized overview, gap map |
| Systematic review | Rigorous evidence synthesis | PRISMA-compliant analysis |
| Meta-analysis | Quantitative evidence pooling | Effect size estimates |
| Scoping review | Mapping breadth of research | Conceptual framework |
| Rapid review | Time-constrained decision | Abbreviated synthesis |
| Competitive analysis | Market positioning | Feature matrix, SWOT |
| User research | Understanding user needs | Personas, journey maps |
Synthesis Techniques
Thematic Analysis
1. Familiarize: read all sources thoroughly 2. Code: label interesting features systematically 3. Theme: collate codes into potential themes 4. Review: check themes against data 5. Define: name and define each theme 6. Write: produce coherent narrative
Evidence Matrix
| Study | Method | Sample | Key Finding | Quality | Relevance |
|---|---|---|---|---|---|
| Smith 2023 | RCT | n=500 | 20% improvement | High | Direct |
| Jones 2022 | Case study | n=1 | Novel approach | Medium | Indirect |
Synthesis by Argument
Group findings by:
- Agreement: Multiple sources converge
- Disagreement: Contradictory findings (explore why)
- Gap: No research on specific sub-question
- Methodological issues: Limitations affecting interpretation
Citation & Reference Management
APA Style (Common for Research)
Book: Author, A. A. (Year). Title. Publisher.
Journal: Author, A. A. (Year). Title. Journal, vol(issue), pages.
Web: Author. (Year, Month Day). Title. Site. URLIEEE Style (Technical/Engineering)
[1] A. Author, "Title of paper," in Title of Book, City, State, Country: Publisher, year, pp. xx-xx.Chicago Style (Humanities/Business)
Author, Title (Place: Publisher, Year), page.Reference Management Tools
| Tool | Best For | Cost |
|---|---|---|
| Zotero | General research | Free |
| Mendeley | Collaboration | Free/Paid |
| EndNote | Institutional | Paid |
| Paperpile | Google Workspace | Paid |
| Citavi | Windows, knowledge management | Paid |
Interview Techniques
SME Interview Structure
1. Context: What are we documenting and why? 2. Walkthrough: Show me how you do X 3. Edge cases: What can go wrong? 4. Audience: Who needs this and what do they know? 5. Validation: Can I follow up with a draft?
Active Listening for Writers
- Paraphrase to confirm understanding
- Ask "Why?" to get rationale, not just steps
- Capture exact terminology (don't translate yet)
- Record with permission, take verbatim notes
Competitive Analysis
Feature Comparison Matrix
| Feature | Us | Competitor A | Competitor B | Notes |
|---|---|---|---|---|
| SSO | Yes | Yes | No | Okta, Azure AD |
| Audit logs | 90 days | 1 year | 30 days | Enterprise need |
| Pricing | $10/user | $15/user | Freemium | We are mid-market |
Analysis Dimensions
- Product features and UX
- Pricing and packaging
- Market positioning and messaging
- Strengths, weaknesses, opportunities, threats
- Customer reviews and sentiment
- Technical architecture (if relevant)
Evidence Hierarchy
| Rank | Source Type | Reliability |
|---|---|---|
| 1 | Systematic reviews, meta-analyses | Highest |
| 2 | Randomized controlled trials | High |
| 3 | Cohort studies | Medium-High |
| 4 | Case-control, case series | Medium |
| 5 | Expert opinion, editorials | Lowest |
For technical research:
- Primary: RFCs, official documentation, source code
- Secondary: Books, peer-reviewed papers, conference talks
- Tertiary: Blog posts, tutorials, Stack Overflow
Bias Detection
| Bias Type | How to Detect | Mitigation |
|---|---|---|
| Publication bias | Positive results overrepresented | Search grey literature |
| Confirmation bias | Seeking sources that agree | Actively search for contradictory evidence |
| Recency bias | Only recent sources | Include seminal older works |
| Authority bias | Assuming expert is always right | Check methodology independently |
| Selection bias | Non-representative sample | Document inclusion/exclusion criteria |
Technical Writing
Document Templates
README Template
# Project Name
One-sentence description of what this project does.
## Installation
pip install package-name
## Quick Start
import package result = package.do_something()
## Documentation
Full docs: [link]
## Contributing
See [CONTRIBUTING.md](CONTRIBUTING.md)
## License
[MIT](LICENSE)API Reference Template
## GET /resource/{id}
Retrieve a single resource by ID.
### Parameters
| Name | Type | Required | Description |
|---|---|---|---|
| id | string | Yes | Unique resource identifier |
### Response{ "id": "abc123", "name": "Example", "created_at": "2024-01-15T10:00:00Z" }
### Errors
| Status | Description |
|---|---|
| 404 | Resource not found |
| 403 | Insufficient permissions |Runbook Template
# [Incident Name] Runbook
## Symptoms
- [ ] Alert: `error_rate > 5%`
- [ ] User reports: [specific symptom]
## Impact Assessment
1. Check dashboard: [link]
2. Identify scope: [affected regions/services]
## Resolution Steps
1. [Step 1 with command]
2. [Step 2 with command]
3. Verify: [how to confirm fix]
## Rollback
If resolution fails:
1. [Rollback command]
2. [Verification]
## Post-Incident
- [ ] Update status page
- [ ] Log in incident trackerStyle Guide Principles
Clarity
- Use simple words: "use" not "utilize"
- Short sentences (<25 words)
- One idea per paragraph
- Front-load important information
Consistency
- Same term for same concept (don't alternate "user" and "customer")
- Same formatting for same elements (all code in backticks)
- Same voice throughout (active preferred for instructions)
Accessibility
- Alt text for all images
- Don't rely on color alone (add labels or patterns)
- Use descriptive link text (not "click here")
- Tables with headers for screen readers
Code Examples
- Complete, copy-paste runnable snippets
- Include expected output as comments
- Show error handling for non-trivial examples
- Update examples when API changes
Information Architecture
Diátaxis Framework
Structure docs by user need, not by feature:
| Type | User Need | Example |
|---|---|---|
| Tutorials | Learning by doing | "Build your first app" |
| How-to guides | Solving a problem | "Deploy to production" |
| Reference | Understanding details | API docs, CLI reference |
| Explanation | Understanding concepts | "How authentication works" |
Doc Site Structure
Getting Started/
Installation
Quick Start
Configuration
Tutorials/
Tutorial 1
Tutorial 2
How-To Guides/
Deploy
Migrate
Troubleshoot
Reference/
API
CLI
Configuration
Explanation/
Architecture
Security
PerformanceReview Rubric
| Criterion | Excellent (5) | Poor (1) |
|---|---|---|
| Accuracy | Verified by SME, code tested | Contains factual errors |
| Clarity | Skimmable, jargon explained | Dense, ambiguous, undefined terms |
| Completeness | Covers all edge cases | Missing critical steps or context |
| Usability | Examples work, links valid | Broken examples, dead links |
| Findability | Properly tagged, cross-linked | Orphan page, no navigation |
Common Anti-Patterns
| Anti-Pattern | Fix |
|---|---|
| "Simply" / "Just" / "Obviously" | Remove; implies the task is trivial |
| Future tense for instructions | Use imperative: "Click Save" not "You will click Save" |
| Wall of text | Break into lists, tables, or steps |
| Missing prerequisites | Add "Before you begin" section |
| Docs that just describe UI | Explain what to do and why |
| Screenshots without alt text | Add descriptive alt text |
| Version-specific without context | State version applicability |
Localization Basics
- Write for translation: avoid idioms, slang, cultural references
- Leave space for text expansion (30% for some languages)
- Don't concatenate strings: "Click" + "Save" → "Click Save"
- Use ICU message format for pluralization and variables
- Provide context for translators in comments
Tools & Frameworks
Markup Languages
| Language | Best For | Example |
|---|---|---|
| Markdown | Developer docs, READMEs, wikis | GitHub, MkDocs, Docusaurus |
| reStructuredText | Python ecosystem, Sphinx | Python docs, Read the Docs |
| AsciiDoc | Enterprise, complex structures | Antora, Asciidoctor |
| DITA | Structured content, reuse | Oxygen, IBM docs |
| XML | Legacy, highly structured | MadCap Flare, FrameMaker |
Docs-as-Code Stack
Static Site Generators
| Tool | Language | Best For |
|---|---|---|
| Docusaurus | React/Node | Open source projects, versioning |
| MkDocs | Python | Simple, fast setup |
| GitBook | SaaS | Collaboration, GitHub sync |
| Hugo | Go | Performance, large sites |
| Gatsby | React | Custom designs, MDX |
| Jekyll | Ruby | GitHub Pages, blogs |
| Antora | Node | Modular docs, multiple repos |
Documentation Platforms
| Platform | Features | Cost |
|---|---|---|
| ReadMe | API docs, developer hubs | Paid |
| Stoplight | API design + docs | Freemium |
| Postman | API testing + documentation | Freemium |
| Swagger UI | OpenAPI rendering | Open source |
| Redoc | OpenAPI rendering | Open source |
Linting & Quality
| Tool | Checks | Integration |
|---|---|---|
| Vale | Style, readability, terminology | CLI, VS Code, CI |
| markdownlint | Markdown formatting | CLI, VS Code |
| proselint | English prose style | CLI |
| write-good | Passive voice, weasel words | CLI, editor |
| alex | Inclusive language | CLI, editor |
Vale configuration example (.vale.ini):
StylesPath = .github/styles
MinAlertLevel = suggestion
Packages = Google, write-good
[*.md]
BasedOnStyles = Vale, Google, write-good
Google.Headings = NODiagram Tools
| Tool | Output | Best For |
|---|---|---|
| Mermaid | Markdown-embedded diagrams | GitHub, docs sites |
| PlantUML | UML diagrams | Architecture docs |
| Draw.io | General diagrams | Visual design |
| Excalidraw | Hand-drawn style | Sketches, wireframes |
| Graphviz | Graph structures | Dependency maps |
Mermaid example:
graph TD
A[User Request] --> B{Cache Hit?}
B -->|Yes| C[Return Cached]
B -->|No| D[Query Database]
D --> E[Cache Result]
E --> F[Return Data]API Documentation
OpenAPI / Swagger
openapi: 3.0.0
info:
title: Example API
version: 1.0.0
paths:
/users/{id}:
get:
summary: Get user by ID
parameters:
- name: id
in: path
required: true
schema:
type: string
responses:
'200':
description: User found
content:
application/json:
schema:
$ref: '#/components/schemas/User'Code Annotation Tools
| Language | Tool | Format |
|---|---|---|
| Python | Sphinx, mkdocstrings | Docstrings |
| JavaScript | JSDoc, TypeDoc | JSDoc comments |
| Java | Javadoc | Javadoc comments |
| Go | godoc | Godoc comments |
| Rust | rustdoc | Markdown comments |
Research Tools
| Tool | Purpose | Cost |
|---|---|---|
| Google Scholar | Academic search | Free |
| Semantic Scholar | AI-powered paper search | Free |
| Zotero | Reference management | Free |
| Notion / Obsidian | Research notes, linking | Freemium |
| Elicit | AI research assistant | Freemium |
| Connected Papers | Citation graph exploration | Freemium |
Version Control for Docs
Git Workflow for Documentation
main (published)
└── develop (staging)
└── feature/docs-update-123Branch naming:
docs/update-readmedocs/api-version-2docs/fix-typo-getting-started
Commit conventions:
docs: update API reference for v2
docs: fix broken links in tutorial
docs: add troubleshooting sectionCollaboration Workflows
Review Process
1. Writer creates PR with documentation changes 2. Automated checks (Vale, link checker, build) 3. SME review for accuracy (request or required) 4. Editor review for style and clarity 5. Merge and deploy (continuous or scheduled)
Feedback Collection
- In-page feedback widget ("Was this helpful?")
- GitHub issues with "documentation" label
- Slack/Discord channel for doc questions
- Quarterly user surveys
Accessibility Tools
| Tool | Check | Integration |
|---|---|---|
| axe DevTools | WCAG compliance | Browser extension |
| WAVE | Accessibility issues | Web, browser |
| Lighthouse | Multiple audits | Chrome, CI |
| pa11y | Automated testing | CLI, CI |
Content Management
| Tool | Type | Best For |
|---|---|---|
| Sanity | Headless CMS | Structured content |
| Contentful | Headless CMS | Enterprise scale |
| Strapi | Open source CMS | Self-hosted |
| Notion | Wiki + database | Team knowledge base |
| Confluence | Enterprise wiki | Atlassian ecosystem |
| GitBook | Docs platform | GitHub-centric |