Waveterm

{
  "service_type": "platform",
  "base_url": "https://www.waveterm.dev",
  "auth_method": "none",
  "auth_config": {},
  "endpoints": [],
  "pricing_model": {
    "type": "free",
    "details": {
      "note": "Waveterm is an open-source application distributed for free under the Apache 2.0 license. There is no paid tier, subscription, or license fee for the software itself. Optional costs may arise from usage of third-party services (AI API providers) that the terminal can integrate with."
    }
  },
  "rate_limits": {},
  "capabilities": [
    "AI assistant integration",
    "Terminal emulation",
    "Shell integration (bash, zsh, fish, etc.)",
    "Command autocomplete and suggestions",
    "Cross-platform support (Linux, macOS, Windows via WSL)",
    "Plugin/extension system",
    "Customizable themes and appearance",
    "Tab and pane management",
    "Session persistence and restore",
    "Integration with editors and IDEs",
    "Clipboard and history management",
    "Keyboard shortcuts and command palette",
    "Remote SSH connections",
    "Scripting and automation",
    "Open-source and self-hosted"
  ],
  "raw_analysis": "Waveterm is an open-source, AI-integrated, cross-platform terminal application designed to improve developer workflows by combining terminal functionality with AI assistance. It is aimed at developers, DevOps engineers, and anyone who frequently works in a command-line environment. The platform is built with a focus on seamless integration of AI capabilities, such as natural language command suggestions and code generation, directly within the terminal interface.\n\nThe project is in active development but has gained significant traction and is considered mature enough for daily use, with a growing community and regular updates. It is distributed under the Apache 2.0 license, making it free to use and modify. There is no public REST API; interactions are primarily through the desktop application and its internal plugin system. The plugin system allows users to extend functionality using JavaScript or TypeScript, though this is not a remote API.\n\nIntegrations include support for popular shells (bash, zsh, fish), the ability to connect to remote servers via SSH, and integration with AI models through API keys (e.g., OpenAI, Anthropic) that users provide. The application can also integrate with local development tools and editors. Pricing is free, with optional costs incurred if users consume third-party AI services. Rate limits are not applicable to the core service, as it runs locally, though any integrated AI APIs will have their own rate limits.\n\nWaveterm's key strengths are its open-source nature, AI integration, cross-platform support, and extensibility. It does not expose a public REST API for programmatic control; instead, it offers a CLI and a plugin API for local automation. As such, it is classified as a platform rather than a service with accessible endpoints."
}

2/3 tests passed

{
  "overall": 68,
  "dimensions": {
    "token_efficiency": 8.0,
    "first_try_success": 7.0,
    "response_parseability": 6.0,
    "error_clarity": 6.5,
    "doc_quality": 7.0,
    "auth_simplicity": 8.0,
    "latency": 9.5,
    "consistency": 8.0
  },
  "pricing_normalized": {
    "model": "free",
    "license": "Apache 2.0",
    "cost": "$0",
    "notes": "Optional third-party AI API costs if user connects external providers; no platform fee."
  },
  "issues": [
    "No llms.txt (404) — agents cannot fetch a machine-readable capability summary",
    "Marketing site is built on Framer, so structured data is limited, though robots.txt is present",
    "No API endpoint exposed; capability info must be inferred from docs/README rather than a programmatic surface",
    "Setup requires local install (Linux/macOS/WSL), so a user can't be productive in a single web session like a SaaS tool"
  ],
  "recommendations": [
    "Publish an /llms.txt with a concise feature/pricing summary to improve agent token efficiency",
    "Provide a public JSON capability or status endpoint for machine-readable access",
    "Include a getting-started guide with concrete time-to-first-command estimates to boost first_try_success",
    "Add structured markdown docs (or exportable docs) so agents can parse feature lists without scraping Framer HTML",
    "Document supported AI backends and required API keys clearly under a 'Requirements' section to improve error_clarity"
  ]
}