Glama MCP Registry Scraper

Scrape the Glama MCP registry — the largest open catalog of Model Context Protocol servers with 23,000+ entries — via the site's public REST API. Returns server metadata, tool schemas, attributes (official, remote-capable), namespaces, source repositories, licenses, and environment-variable schemas. Pairs cleanly with the Smithery MCP scraper for full ecosystem coverage.

---

Glama MCP Scraper Features

• Crawls Glama's full MCP server catalog — 23,000+ servers indexed
• Three modes: full catalog dump, search-by-query, or single server detail
• Returns tool schemas exposed by each server, not just names
• Captures Glama attributes (author:official, hosting:remote-capable, etc.) for filtering
• Includes source repository URL, SPDX license, namespace, and slug per server
• Returns environment-variable JSON schemas as a JSON string, ready to drop into a config UI
• Cursor-paginated through the public API — runs to completion without manual pagination
• No API key, no proxy, no captcha

---

Who Uses MCP Registry Data?

• MCP client developers — Build a server picker UI that lists every available MCP server with metadata
• AI agent platforms — Pre-index the MCP ecosystem and surface relevant servers to users at runtime
• Security researchers — Audit server attributes, source repos, and license distribution across the open MCP catalog
• MCP framework maintainers — Monitor namespace growth, attribute distribution, and tool-schema patterns
• Developer-tool indexers — Add MCP servers to broader AI-tooling search platforms with first-class metadata
• Internal AI ops teams — Maintain an allow-list of vetted MCP servers with full schemas for self-hosted deployment

---

How the Glama MCP Scraper Works

1. Pick a mode — catalog for the full registry, search for a query-filtered slice, or server_detail for a single server by ID
2. The scraper hits the Glama REST API — Cursor-paginates 100 servers per page through the public endpoint
3. Optionally filters by attribute or primary language — Pass author:official or Python to narrow the result set
4. Returns one record per server with metadata, tools, attributes, repo URL, license, and env-var schema

The API is public and unauthenticated. The scraper observes a 200ms delay between requests — polite enough not to draw attention, fast enough to walk the full catalog without setting up camp.

---

Input

{
  "mode": "catalog",
  "maxItems": 500
}

Field	Type	Default	Description
mode	string	catalog	One of catalog, search, server_detail.
query	string	""	Search query for search mode (server name, description, or tags).
serverId	string	""	Glama server ID for server_detail mode.
attributeFilter	string	""	Filter by Glama attribute (e.g. author:official, hosting:remote-capable).
languageFilter	string	""	Filter by primary language (Python, TypeScript, Go, Rust).
maxItems	integer	500	Maximum servers to return. Set to 0 for unlimited.
proxyConfiguration	object	none	Proxy settings. Not required — the public API accepts datacenter IPs.

Search by query

{
  "mode": "search",
  "query": "github",
  "maxItems": 100
}

Single server detail

{
  "mode": "server_detail",
  "serverId": "glama-server-id-here"
}

---

Glama MCP Scraper Output Fields

{
  "server_id": "abc123",
  "name": "GitHub MCP Server",
  "description": "Read-only MCP server for GitHub repositories, issues, and pull requests.",
  "attributes": ["author:official", "hosting:remote-capable"],
  "tools": ["search_repos|Search GitHub repositories", "get_issue|Fetch an issue by number"],
  "resources": [],
  "prompts": [],
  "install_command": null,
  "source_repo_url": "https://github.com/example/github-mcp",
  "homepage_url": null,
  "license": "MIT",
  "language": null,
  "star_count": null,
  "namespace": "github",
  "slug": "github-mcp",
  "env_schema_json": "{\"type\":\"object\",\"properties\":{\"GITHUB_TOKEN\":{\"type\":\"string\"}}}",
  "server_url": "https://glama.ai/mcp/servers/abc123"
}

Field	Type	Description
server_id	string	Glama server ID.
name	string	Server name.
description	string	Server description.
attributes	array	Glama attributes (e.g. hosting:remote-capable, author:official).
tools	array	Tools exposed by this server, formatted as `name
resources	array	Resources exposed by this server.
prompts	array	Prompts exposed by this server.
install_command	string	Install command (when published by the server author).
source_repo_url	string	Source repository URL.
homepage_url	string	Homepage URL.
license	string	SPDX license name.
language	string	Primary language inferred from the repository.
star_count	integer	GitHub star count, when linked.
namespace	string	Glama namespace (author or org).
slug	string	Glama server slug.
env_schema_json	string	Environment-variable JSON schema as a JSON string.
server_url	string	Glama server page URL.

---

FAQ

How do I scrape the Glama MCP registry?

Glama MCP Scraper hits the public REST API at glama.ai/api/mcp/v1/servers and cursor-paginates through the catalog. No auth, no proxy, no captcha. Pick a mode, set maxItems, and run.

How many MCP servers can I get?

Glama MCP Scraper covers 23,000+ servers — every entry in the Glama registry at crawl time. Set maxItems to 0 to pull the full catalog or pass a number to sample.

Can I filter to only official or remote-capable servers?

Glama MCP Scraper accepts an attributeFilter like author:official or hosting:remote-capable. The filter is applied during the crawl, so you don't pay for records you'll discard.

How much does the Glama MCP Scraper cost to run?

Glama MCP Scraper is priced per record returned via the pay-per-event model. A full-catalog run lands well under the cost of building and maintaining your own registry crawler.

Does this need proxies?

Glama MCP Scraper does not need proxies. The public API accepts datacenter IPs without rate-limiting beyond the polite 200ms delay the scraper observes between requests.

---

Need More Features?

Need additional fields, multi-source MCP registry merging (Smithery + Glama + others), or scheduled updates? File an issue or get in touch.

Why Use the Glama MCP Scraper?

• Largest open registry — Glama indexes 23K+ servers, well ahead of every other MCP catalog
• Real tool schemas — Returns the tools each server exposes with names and descriptions, not just a count. Useful when you want to surface capability summaries.
• Clean JSON — Normalized records with consistent field names, env-var schemas pre-stringified, ready to drop into a config UI or registry mirror