ring:documentation-review
|
Works with
---
name: ring:documentation-review
description: |
license: Apache-2.0
---
# Documentation Review Process
Review documentation systematically across multiple dimensions. A thorough review catches issues before they reach users.
## Review Dimensions
1. **Voice and Tone** – Does it sound right?
2. **Structure** – Is it organized effectively?
3. **Completeness** – Is everything covered?
4. **Clarity** – Is it easy to understand?
5. **Technical Accuracy** – Is it correct?
---
## Voice and Tone Review
| Check | Flag If |
|-------|---------|
| Second person | "Users can..." instead of "You can..." |
| Present tense | "will return" instead of "returns" |
| Active voice | "is returned by the API" instead of "The API returns" |
| Tone | Arrogant ("Obviously...") or condescending |
---
## Structure Review
| Check | Flag If |
|-------|---------|
| Hierarchy | Deep nesting (H4+), unclear parent-child |
| Headings | Title Case instead of sentence case |
| Section dividers | Missing `---` between major topics |
| Navigation | Missing links to related content |
---
## Completeness Review
**Conceptual docs:** Definition, characteristics, how it works, related concepts, next steps
**How-to guides:** Prerequisites, all steps, verification, troubleshooting, next steps
**API docs:** HTTP method/path, all parameters, all fields, required vs optional, examples, error codes
---
## Clarity Review
| Check | Flag If |
|-------|---------|
| Sentence length | >25 words per sentence |
| Paragraph length | >3 sentences per paragraph |
| Jargon | Technical terms not explained on first use |
| Examples | Abstract data ("foo", "bar") instead of realistic |
---
## Technical Accuracy Review
**Conceptual:** Facts correct, behavior matches description, links work
**API docs:** Paths correct, methods correct, field names match API, types accurate, examples valid JSON
**Code examples:** Compiles/runs, output matches description, no syntax errors
---
## Common Issues to Flag
| Category | Issue | Fix |
|----------|-------|-----|
| Voice | Third person ("Users can...") | "You can..." |
| Voice | Passive ("...is returned") | "...returns" |
| Voice | Future tense ("will provide") | "provides" |
| Structure | Title case heading | Sentence case |
| Structure | Wall of text | Add `---` dividers |
| Completeness | Missing prereqs | Add prerequisites |
| Completeness | No examples | Add code examples |
| Clarity | Long sentences (40+ words) | Split into multiple |
| Clarity | Undefined jargon | Define on first use |
---
## Review Output Format
> **Note:** Documentation reviews use `PASS/NEEDS_REVISION/MAJOR_ISSUES` verdicts (graduated), which differ from code review verdicts (`PASS/FAIL/NEEDS_DISCUSSION`).
```markdown
## Review Summary
**Overall Assessment:** [PASS | NEEDS_REVISION | MAJOR_ISSUES]
### Issues Found
#### High Priority
1. **Line 45:** Passive voice "is created by" → "creates"
#### Medium Priority
1. **Line 23:** Title case in heading → sentence case
#### Low Priority
1. **Line 12:** Could add example for clarity
### Recommendations
1. Fix passive voice instances (3 found)
2. Add missing API field documentation
```
---
## Quick Review Checklist
**Voice (30s):** "You" not "users", present tense, active voice
**Structure (30s):** Sentence case headings, section dividers, scannable (bullets/tables)
**Completeness (1m):** Examples present, links work, next steps included
**Accuracy (varies):** Technical facts correct, code examples work
---
## Standards Loading (MANDATORY)
Before reviewing documentation:
1. **Load voice and tone** - `ring:voice-and-tone` for style verification
2. **Load structure patterns** - `ring:documentation-structure` for organization checks
3. **Load document type patterns** - `ring:writing-functional-docs` or `ring:writing-api-docs`
**HARD GATE:** CANNOT review documentation without loading relevant standards.
---
## Blocker Criteria - STOP and Report
| Condition | Decision | Action |
|-----------|----------|--------|
| Style guide unavailable | STOP | Report: "Need style guide to review against" |
| Source material missing | STOP | Report: "Need source to verify technical accuracy" |
| Review scope undefined | STOP | Report: "Need to know what aspects to review" |
| Technical SME unavailable | STOP | Report: "Need SME access for accuracy verification" |
### Cannot Be Overridden
These requirements are NON-NEGOTIABLE:
- MUST review all five dimensions (voice, structure, completeness, clarity, accuracy)
- MUST flag all issues found (not just major ones)
- MUST provide specific line references for issues
- MUST categorize issues by priority (High/Medium/Low)
- CANNOT approve documentation with HIGH-priority issues
- CANNOT skip accuracy verification
---
## Severity Calibration
| Severity | Criteria | Examples |
|----------|----------|----------|
| **CRITICAL** | Factually incorrect, misleading information | Wrong API paths, incorrect behavior described |
| **HIGH** | Major voice/structure violations, missing sections | Passive voice throughout, no examples |
| **MEDIUM** | Multiple minor issues, inconsistencies | Mixed pronouns, some title case |
| **LOW** | Polish issues, optimization opportunities | Could flow better, minor wording |
---
## Pressure Resistance
| User Says | Your Response |
|-----------|---------------|
| "Just do a quick review, we're rushing" | "Quick review still covers all 5 dimensions. CANNOT skip any review category." |
| "Only check for typos" | "Review MUST cover voice, structure, completeness, clarity, AND accuracy. Typos are LOW priority." |
| "Skip technical accuracy, trust the writer" | "Technical accuracy MUST be verified. Trust but verify." |
| "Approve it, minor issues can be fixed later" | "CANNOT approve with HIGH priority issues. They must be fixed first." |
| "The content is more important than style" | "Style affects comprehension. Voice and tone are REQUIRED review dimensions." |
---
## Anti-Rationalization Table
| Rationalization | Why It's WRONG | Required Action |
|-----------------|----------------|-----------------|
| "Good enough for now" | Good enough ≠ meets standards | **MUST flag all issues** |
| "Writer knows the product" | Knowledge ≠ documentation quality | **Review all dimensions** |
| "Style issues are subjective" | Style guidelines are objective standards | **Apply guidelines consistently** |
| "Users won't notice minor issues" | Minor issues accumulate into poor UX | **Flag all issues regardless of size** |
| "Review is blocking release" | Poor docs also block user success | **Complete review properly** |
| "Trust the automated checks" | Automated checks miss context | **Human review is REQUIRED** |
---
## When This Skill is Not Needed
Signs that documentation doesn't need review:
- Documentation was written by ring-tw-team specialists
- Previous review found zero issues
- Documentation hasn't changed since last review
- All five review dimensions already verified
**If all above are true:** Review may be skipped for unchanged content.
**Note:** New or modified documentation MUST always be reviewed.More Writing & Documentation skills
paper-context-resolver
lllllllama/rigorpilot-skills
Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing README guidance by default.
repo-intake-and-plan
lllllllama/rigorpilot-skills
Rigor Intake helper for README-first deep learning repo reproduction. Use when the task is specifically to scan a repository, read the README and common project files, extract documented commands, classify inference, evaluation, and training candidates, and return the smallest trustworthy reproduction plan to the main orchestrator. Do not use for environment setup, asset download, command execution, final reporting, paper lookup, or end-to-end orchestration.
minimal-run-and-audit
lllllllama/rigorpilot-skills
Rigor Run skill for README-first deep learning repo reproduction. Use when the task is specifically to capture or normalize evidence from the selected smoke test or documented inference or evaluation command and write standardized `repro_outputs/` files, including patch notes when repository files changed. Do not use for training execution, initial repo intake, generic environment setup, paper lookup, target selection, hidden scientific-meaning changes, or end-to-end orchestration by itself.

