{
  "transport": "stdio",
  "command": "./mcp-servers/navi-web/.venv/bin/python",
  "args": [
    "-m",
    "app.mcp_server"
  ],
  "cwd": "./mcp-servers/navi-web",
  "env": {
    "NAVI_WEB_MCP_TRANSPORT": "stdio",
    "NAVI_WEB_SEARXNG_URL": "${SEARXNG_URL:-}"
  },
  "groups": {
    "search": [
      "web_search"
    ],
    "browse": [
      "web_view"
    ],
    "request": [
      "http_request"
    ]
  },
  "instructions": "Navi Web MCP server provides web search, browsing, and raw HTTP tools.\n\nUse it when the task involves:\n- searching the web for current info, docs, or real-time data;\n- opening a URL in a browser to read human-readable content;\n- making REST API calls, webhooks, or raw HTTP requests.\n\nWorkflow:\n1. search — find relevant pages or facts.\n2. view — open promising URLs to read full content.\n3. request — call APIs or services requiring headers/auth.\n\nBatching: web_search takes one query or up to 3, web_view one URL or up to 3 — several run in parallel, so the call costs the slowest instead of the sum. Use it when you already know the lookups are independent: after a search, open the 2-3 pages you need in one web_view call; when a request needs several unrelated facts, ask the searches together. A failed item in a batch is named in the output rather than dropping the whole call. Do not batch a follow-up that depends on what the previous answer said — that is a sequence, not a batch.\n\nAll three tools are stateless and work with public URLs.\nNo session_id or filesystem paths are required.\n\nABSOLUTE RULE — NEVER bypass MCP tools:\nYou MUST NOT use filesystem, terminal, code_exec, or any direct file access to read or write web content. Use only the MCP tools listed above.",
  "summary": "Web search, page browsing and raw HTTP; up to 3 queries or URLs in one call run in parallel, so batch independent lookups. Never fetch web content with filesystem, terminal or code_exec — these are the only valid interface. Details: tool_manual(\"navi-web\")."
}