Zeal

```json
{
  "service_type": "platform",
  "base_url": "https://zealdocs.org",
  "auth_method": "none",
  "auth_config": {},
  "endpoints": [],
  "pricing_model": {
    "type": "free",
    "details": {
      "description": "Free and open-source (GPL-3.0) desktop application. No account or paid tier.",
      "source": "https://zealdocs.org"
    }
  },
  "rate_limits": {},
  "capabilities": [
    "offline documentation browsing",
    "docset management (browse, search, install)",
    "full-text search across docsets",
    "docset repository integration (downloadable documentation sets)",
    "keyboard-driven navigation",
    "syntax highlighting in code samples",
    "cross-platform desktop app (Windows, macOS, Linux)",
    "no internet connection required to view docs",
    "integration with editors/IDEs via external lookups (capability of the broader ecosystem)",
    "plugin/extension-style docset additions (custom docsets in Dash-compatible format)"
  ],
  "raw_analysis": "Zeal is a free, open-source, offline documentation browser for developers, directly inspired by the macOS Dash app. It runs as a native desktop application on Windows, macOS, and Linux. Its core value proposition is giving developers instant, offline access to API and language documentation without relying on the web or a specific IDE.\n\nWhat it does: Zeal downloads 'docsets' (bundled documentation in Dash-compatible format) from a central repository and lets users browse and full-text search them locally. Users can manage installed docsets, add custom docset sources, and quickly look up symbols/terms while coding.\n\nWho it's for: Software developers, especially those working in environments with limited/no internet connectivity, or those who want fast local lookups over slower web docs. It appeals to users of editors like Vim, Emacs, Sublime Text, and VS Code via companion integrations, though those integrations live in the editor ecosystem rather than in Zeal itself.\n\nMaturity: Zeal is a well-established, widely used open-source project. It is distributed through official channels and commonly packaged in Linux distros. Development has historically been community-driven with periodic releases; it is considered stable and mature but not rapidly evolving.\n\nIntegrations: Zeal does not expose a public REST API. Its integration surface is: (1) the Dash docset format (feeds of downloadable documentation), and (2) the ability to be invoked from editors/IDEs via local commands or community plugins. There is no hosted service, no authentication, no cloud component by default \u2014 it is a local desktop tool.\n\nPlatform/tooling caveat: Because this is a client-side desktop application rather than a hosted SaaS or API service, there are no HTTP endpoints, no API keys, no rate limits, and no subscription pricing. Programmatic/API access would have to go through the underlying docset feed format (an XML 'feed' listing docsets plus tarballs), not through Zeal itself."
}
```

2/3 tests passed

{
  "overall": 68,
  "dimensions": {
    "token_efficiency": 8.5,
    "first_try_success": 7.5,
    "response_parseability": 5.0,
    "error_clarity": 6.0,
    "doc_quality": 7.0,
    "auth_simplicity": 10.0,
    "latency": 9.5,
    "consistency": 8.0
  },
  "pricing_normalized": {
    "model": "free",
    "open_source": true,
    "license": "GPL-3.0",
    "paid_tiers": 0,
    "notes": "Fully free, no account required, desktop app only"
  },
  "issues": [
    "Desktop-only application — no web API, no cloud service, no programmatic access for agents",
    "No llms.txt (404) — no machine-readable context for AI agents",
    "No structured API/export — response_parseability limited to HTML scraping",
    "Docsets are download-and-hold (offline model) — not a live queryable endpoint",
    "Requires human install of a desktop binary; agent cannot directly fetch/use documentation"
  ],
  "recommendations": [
    "Add an llms.txt to describe project, docsets, and capabilities for AI agent discovery",
    "Expose a public JSON catalog of available docsets (name, version, URL) for programmatic consumption",
    "Consider a web search endpoint or static JSON index of docset contents for agents",
    "Improve docs with a machine-readable features/limits section",
    "Recommend only when the user wants a local, offline documentation browser — not for API-driven workflows"
  ]
}