From 287f93bb4d5ec5aa485b2751514ac4651fe48833 Mon Sep 17 00:00:00 2001 From: Grady Dillon Date: Thu, 23 Jul 2026 20:20:01 -0400 Subject: [PATCH] Document the search tool in the README MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The server exposes a `search` tool (web search → extraction in one call), but the README's tool table, the remote-connector capability line, and the intro all predated it. Add `search` to each, and note the shared params (`format`, compliance overrides, `retry`, `max_pages`) the tools accept. Docs only. Co-Authored-By: Claude Opus 4.8 --- README.md | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index f4f9828..1161748 100644 --- a/README.md +++ b/README.md @@ -2,8 +2,9 @@ The official [Model Context Protocol](https://modelcontextprotocol.io) server for [WellMarked.io](https://wellmarked.io) — give your AI agents the ability to turn -any URL into clean Markdown, bulk-extract batches of pages, and crawl entire -sites, all from inside Claude Desktop, Claude Code, Cursor, or any MCP host. +any URL into clean Markdown, bulk-extract batches of pages, crawl entire sites, +and search the web, all from inside Claude Desktop, Claude Code, Cursor, or any +MCP host. It's a thin adapter over the official [`wellmarked`](https://www.npmjs.com/package/wellmarked) JavaScript SDK, so it inherits the SDK's auth, typed errors, retry/back-off @@ -16,10 +17,17 @@ hints, and polymorphic job polling. | `extract` | Fetch one URL and return its main content as clean Markdown + metadata. | | `bulk` | Submit many URLs for concurrent extraction (Pro+). Blocks for results by default. | | `crawl` | Crawl a site BFS from a root URL to a given depth (Pro+). Returns an async job. | +| `search` | Search the web and return each result page as clean Markdown, in one call (Pro+). | | `get_job` | Poll a bulk/crawl job once by id. | | `wait_for_job` | Block until a job finishes, then return all results. | | `get_usage` | Report current billing-period quota (plan, used, limit, remaining). | +`extract`, `bulk`, `crawl`, and `search` all accept a `format` +(`markdown` / `json` / `chunks` / `html` / `links`) and per-request compliance +overrides (`allow_domains` / `deny_patterns` / `respect_robots`); `extract`, +`bulk`, and `crawl` also accept a `retry` count for target timeouts, and +`crawl` a `max_pages` budget. + > **Not exposed as tools:** API-key rotation (`/keys/rotate`) and webhook-secret > rotation (`/webhook/rotate`). Both are destructive and irreversible — the old > credential dies the instant the call returns, with no recovery flow — which is @@ -47,8 +55,8 @@ https://mcp.wellmarked.io/mcp Your host discovers the authorization server automatically (via `/.well-known/oauth-protected-resource`), walks you through signing in to WellMarked and approving the connection, and receives a scoped, expiring token — -no key to paste. The connector can `extract`, `bulk`, and `crawl`; it cannot -mint or rotate credentials. +no key to paste. The connector can `extract`, `bulk`, `crawl`, and `search`; it +cannot mint or rotate credentials. In **Claude.ai**: Settings → Connectors → Add custom connector → paste the URL above → follow the sign-in and consent prompts.