{
  "markdown": "# comber-mcp\r\n\r\n<!-- mcp-name: io.codecomber/comber -->\r\n\r\nAn [MCP](https://modelcontextprotocol.io) server for [Comber](https://codecomber.io) — the\r\nevery-click QA agent. Point your editor's agent at a preview URL and it walks every inch of the\r\napp, then reports the broken clicks, dead ends, console errors, and broken links right inline.\r\n\r\nOne server, every MCP client: **Cursor, Claude Desktop, Claude Code, Windsurf, Zed.**\r\n\r\n## Tools\r\n\r\n| Tool | What it does |\r\n| --- | --- |\r\n| `comber_check` | Crawl a web app / preview URL and report findings. Read-only by default. |\r\n| `comber_api_check` | Exercise an OpenAPI spec, GraphQL endpoint, or recorded HAR. |\r\n| `comber_report` | Pull the full findings list from a prior run. |\r\n\r\nChecks are **read-only by default** — Comber does not submit forms or trigger mutations unless you\r\nexplicitly pass `allowWrites`. The optional LLM finding-judge only runs when you have your own\r\n`ANTHROPIC_API_KEY` in the environment; nothing is billed to Comber.\r\n\r\n## Install\r\n\r\nThe server runs straight from npm — no clone, no build.\r\n\r\n### Cursor\r\n\r\n`Settings → MCP → Add new server`, or drop this in `.cursor/mcp.json` (project) or\r\n`~/.cursor/mcp.json` (global):\r\n\r\n```json\r\n{\r\n  \"mcpServers\": {\r\n    \"comber\": {\r\n      \"command\": \"npx\",\r\n      \"args\": [\"-y\", \"comber-mcp\"],\r\n      \"env\": { \"ANTHROPIC_API_KEY\": \"\" }\r\n    }\r\n  }\r\n}\r\n```\r\n\r\nLeave `ANTHROPIC_API_KEY` empty for heuristic-only checks, or set it to enable the LLM judge.\r\n\r\n### Claude Code\r\n\r\n```bash\r\nclaude mcp add comber -- npx -y comber-mcp\r\n```\r\n\r\n### Claude Desktop\r\n\r\nIn `claude_desktop_config.json`:\r\n\r\n```json\r\n{\r\n  \"mcpServers\": {\r\n    \"comber\": { \"command\": \"npx\", \"args\": [\"-y\", \"comber-mcp\"] }\r\n  }\r\n}\r\n```\r\n\r\n### Windsurf\r\n\r\nIn `~/.codeium/windsurf/mcp_config.json` — same `mcpServers` shape as above.\r\n\r\n### Zed\r\n\r\nIn `settings.json`:\r\n\r\n```json\r\n{\r\n  \"context_servers\": {\r\n    \"comber\": { \"command\": { \"path\": \"npx\", \"args\": [\"-y\", \"comber-mcp\"] } }\r\n  }\r\n}\r\n```\r\n\r\n## Usage\r\n\r\nOnce installed, just ask your agent:\r\n\r\n> Run comber on `https://my-branch.preview.app` and fix anything that fails.\r\n\r\nThe agent calls `comber_check`, gets a coverage + findings summary back, and can call\r\n`comber_report` for the full detail before proposing fixes.\r\n\r\nGate a merge:\r\n\r\n> Run comber with strict on the preview and block if there are any fail findings.\r\n\r\n## Notes\r\n\r\n- The first web check may prompt Playwright to download a browser; run\r\n  `npx playwright install chromium` once if you hit a launch error.\r\n- Runaway-cost backstops (`maxStates`, `maxActions`) are exposed as tool inputs.\r\n- Full engine docs: <https://codecomber.io>.",
  "bytes": 2655,
  "sha": "a30ac0bfe7defb4e813a122d2e71b4c6ebd1cdb25165cc183c2452ef57f6fef1",
  "repo_slug": "",
  "fonte": "npm",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_codecomber_comber_9cc68e5f/readme"
}