web2md-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| WEB2MD_API_KEY | No | Optional. Only needed for the library and skill tools; convert_url and convert_file work without it. Get one free at web2md.org. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| convert_urlA | Convert a public webpage to clean Markdown. START HERE for any URL — it needs no browser, no extension, and no setup. Returns Markdown plus metadata (title, word count, reading time). If the page needs your login session or renders only after JavaScript runs (Reddit, X, paywalled articles, internal dashboards), this returns a login wall or an empty shell — switch to agent_convert in that case. |
| convert_fileA | Convert a local document file to clean Markdown. Supports Word (.docx/.doc), PowerPoint (.pptx/.ppt), Excel (.xlsx/.xls), OpenDocument (.odt/.ods/.odp), RTF, EPUB, CSV, and PDF. Tables, headings, lists and footnotes are preserved. Fast, fully local — the file never leaves the machine. |
| semantic_searchA | [deprecated: use search_library/get_library_items] Search your saved conversions using natural language. Works across languages. |
| get_conversionA | [deprecated: use search_library/get_library_items] Retrieve the full Markdown content of a previously saved conversion by its ID. |
| bridge_convert_urlA | FALLBACK for agent_convert when the native host is not installed. Same capability — reads the page through the user's signed-in Chrome — but it works by opening a browser tab, so it is visible and takes focus. Prefer agent_convert whenever the native host is available; use this one if agent_convert reports the host is missing, or on a machine where you cannot install it. Needs only the Web2MD extension installed and signed in; the extension is found automatically via the API key. |
| agent_convertA | Convert one URL through the user's own signed-in Chrome, reading the page exactly as they see it. USE THIS when convert_url hits a login wall or returns an empty shell — Reddit, X, paywalled articles, internal wikis, dashboards. Runs silently in the background over a local connection; it does not open or steal focus from any tab. Requires Chrome running with the Web2MD extension and the native host installed (npx -p web2md-mcp-server web2md-mcp-setup). For several URLs at once use agent_batch_convert; if the native host is not installed use bridge_convert_url. |
| agent_batch_convertA | The multi-URL form of agent_convert: up to 50 URLs in one call, each read through the user's signed-in Chrome. USE THIS instead of calling agent_convert in a loop — one call, one round trip, and each result carries its own token count. URLs are processed one after another, so a long list takes a while. Same requirements as agent_convert (Chrome + extension + native host). |
| list_skillsA | List the skills saved in your Web2MD cloud account. Returns a compact list (id, name, description). Use get_skill to fetch the full SKILL.md body of one. |
| get_skillA | Fetch the full content of a single Web2MD skill by id (or name). Returns its name, description, and complete SKILL.md body. Use list_skills first to find ids. |
| search_libraryA | Search your Web2MD library (saved conversions) by keyword (matches title and full page content) and/or tag, newest first. Returns item metadata: id, title, url, date, tags. Use get_library_items to fetch full markdown. |
| get_library_itemsA | Fetch full markdown content for library items by id (max 20 per call). Use search_library first to find ids. Responses can be very large — for long pages, batch 5 or fewer ids per call. |
| sync_libraryA | Sync your Web2MD library to a local directory as markdown files (organized by tag, with frontmatter). Incremental by default; keeps state in /.web2md-sync.json. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
The four conversion tools (convert_url, agent_convert, bridge_convert_url, agent_batch_convert) all target the same action and rely on descriptions rather than distinct purposes to separate them; agent_convert vs bridge_convert_url is purely a mechanism fallback, which is easy to misselect. Deprecated get_conversion and semantic_search still sit alongside their replacements (get_library_items, search_library), creating redundant overlapping pairs. The descriptions are unusually explicit about when to use which, which rescues the set from a lower score.
Nearly all tools follow snake_case verb_noun or prefix_noun patterns (get_library_items, list_skills, search_library, sync_library, convert_file). Deviations are minor: the agent_/bridge_ mechanism prefixes and the adjective-first semantic_search break the verb pattern slightly. Still highly readable and predictable overall.
Twelve tools is well within a healthy range for a conversion-plus-library server, and each of the non-deprecated tools maps to a real operation. Two of the twelve are explicitly deprecated and only earn their place as legacy aliases, which is mild dead weight.
Conversion coverage is strong (public URL, authenticated browser, batch, local files) and read paths for library and skills exist. However, there is no way to delete or manage library items, and skills are read-only (no create/update/delete), leaving the lifecycle half-covered. The presence of two deprecated tools also suggests the surface has drifted rather than been consolidated.