> ## Documentation Index
> Fetch the complete documentation index at: https://www.doc-reviewer.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Doc Reviewer: LLM-powered documentation quality analysis

> Doc Reviewer helps technical writers evaluate the completeness and quality of instructions in technical documentation using configurable LLM-powered criteria.

Doc Reviewer is a desktop application that automatically scans your technical documentation — PDFs, Word documents, Markdown files, and web pages — detects instruction sections, and evaluates each one against a configurable set of quality criteria using large language models. Every instruction gets a color-coded result (green, yellow, orange, or red) along with specific, actionable recommendations to help you improve your documentation.

<CardGroup cols={2}>
  <Card title="Installation" icon="download" href="/installation">
    Download and set up Doc Reviewer on Windows, connect your LLM, and run your first evaluation.
  </Card>

  <Card title="Quick Start" icon="bolt" href="/quickstart">
    Upload your first document and get evaluation results in minutes.
  </Card>

  <Card title="Core Concepts" icon="book-open" href="/concepts/projects">
    Understand projects, documents, evaluation results, and criteria sets.
  </Card>

  <Card title="Configuration" icon="sliders" href="/configuration/llm-models">
    Connect OpenAI, Anthropic, Ollama, or any OpenAI-compatible model.
  </Card>
</CardGroup>

## How it works

<Steps>
  <Step title="Install and launch">
    Download the `doc-reviewer.exe` and run it. The app starts a local server and opens in your browser automatically.
  </Step>

  <Step title="Connect an LLM">
    Go to **Settings → Models** and add your LLM API key or configure a local model via Ollama.
  </Step>

  <Step title="Create a project and upload documents">
    Create a project for your product, then upload PDF, DOCX, Markdown, or TXT files — or load pages directly from a URL.
  </Step>

  <Step title="Run evaluation and review results">
    Click **Evaluate** to analyze all instruction sections. Review color-coded results and recommendations, then export to XLS.
  </Step>
</Steps>

## Key features

<CardGroup cols={2}>
  <Card title="Multi-format document support" icon="file-lines">
    Parse and evaluate PDF, DOCX, Markdown, and TXT files. Load web pages by URL using a headless browser.
  </Card>

  <Card title="Automatic instruction detection" icon="magnifying-glass">
    Doc Reviewer automatically identifies instruction sections in your documents using morphological analysis — no manual tagging required.
  </Card>

  <Card title="Configurable evaluation criteria" icon="list-check">
    Use the built-in criteria set or create your own. Edit criteria directly in the UI using Markdown format.
  </Card>

  <Card title="Version comparison with snapshots" icon="code-compare">
    Save evaluation results as snapshots and compare them across document versions to track quality improvements over time.
  </Card>

  <Card title="Product context for better accuracy" icon="brain">
    Attach a product context to each project so the LLM understands your product's terminology and audience when evaluating instructions.
  </Card>

  <Card title="Local and private" icon="lock">
    All data stays on your machine. Doc Reviewer uses a local SQLite database and runs entirely offline except for LLM API calls.
  </Card>
</CardGroup>