Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcpcatalogs search for a Postgres MCP server"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP server that lets your AI assistant query mcpcatalogs.com — the independent bilingual directory of MCP servers — directly inside the chat. Search, compare, top-list, and dive into full server details without leaving your IDE or Claude Desktop.
The directory tracks 1,900+ MCP servers, refreshed daily with composite scores combining AI evaluation, GitHub signals, and real-world Smithery usage.
Tool | What it does |
| Free-text search by topic, optional category narrowing |
| Full record: install, FAQ, "when to choose / when NOT", alternatives |
| Side-by-side comparison between two servers (stars, usage, score, decision-guide) |
| Ranked list by composite quality / GitHub stars / real Smithery usage |
All four return clean markdown — drop-in citation source for LLM answers.
Related MCP server: agentalmanac-mcp
Edit your Claude Desktop config:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json
Add:
{
"mcpServers": {
"mcpcatalogs": {
"command": "npx",
"args": ["-y", "mcpcatalogs"]
}
}
}Restart Claude Desktop. You should see four new tools in the tools menu.
In Cursor → Settings → Features → MCP:
{
"mcpcatalogs": {
"command": "npx",
"args": ["-y", "mcpcatalogs"]
}
}Same shape — point the client at npx -y mcpcatalogs.
If your client supports remote MCP servers over HTTP:
{
"mcpServers": {
"mcpcatalogs": {
"url": "https://mcpcatalogs.com/mcp"
}
}
}After installing, ask your AI assistant things like:
"Search the MCP directory for a Postgres server and rank by real usage."
"Compare modelcontextprotocol-server-postgres and supabase-supabase-mcp."
"What's the top browser-automation MCP server right now?"
"Show me the full detail for antvis-mcp-server-chart, including when I should NOT choose it."
The directory refreshes daily from GitHub + Smithery. Each tool call hits a live database, so you always get the latest score, last-commit date, and usage numbers.
git clone https://github.com/mcpcatalogs/mcpcatalogs
cd mcpcatalogs
npm install
npm run build
node dist/index.jsTo point at a custom backend (e.g. a fork), set environment variables:
export MCPCATALOGS_SUPABASE_URL=https://your-project.supabase.co
export MCPCATALOGS_SUPABASE_ANON_KEY=your-anon-keyBug reports and PRs welcome. The directory itself (scraping, scoring, AI evaluation) lives in a separate private repo — this package is just the MCP-protocol interface to it.
MIT. See LICENSE.
Side-by-side comparison of two MCP servers across stars, real usage, composite score, official status, categories, language, last commit, and decision-guide content.
| Name | Required | Description | Default |
|---|---|---|---|
| slug_a | Yes | First server slug, e.g. "modelcontextprotocol-server-postgres". | |
| slug_b | Yes | Second server slug. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden. It lists the fields compared, giving some insight into the output, but does not disclose whether the tool is read-only, data freshness, or any side effects. This is adequate but not thorough.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that efficiently lists the comparison fields without any wasted words. It is front-loaded with the core action and resource.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple comparison tool with two required parameters, the description sufficiently covers the purpose and scope. It does not explain the output format in detail or error handling, but given the lack of output schema, it is reasonably complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, and the description's mention of slugs matches the schema. The description does not add new parameter semantics beyond what the schema already provides (e.g., format examples). Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Side-by-side comparison of two MCP servers' and enumerates specific comparison fields (stars, usage, score, etc.), which distinguishes it from siblings like get_server_detail (single server) and search_mcp_servers (searching).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies the tool is used for comparing two servers, but does not explicitly state when to use it instead of alternatives (e.g., get_server_detail for single server info), nor does it mention when not to use it or any prerequisites.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get the full record for a single MCP server: summary, install snippet, use cases, FAQ, 'when to choose / when NOT', and links.
| Name | Required | Description | Default |
|---|---|---|---|
| slug | Yes | Server slug, normally "{author}-{repo}". Example: "modelcontextprotocol-server-postgres". You can find slugs via search_mcp_servers or list_top first. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, but the description accurately characterizes this as a read operation ('Get'). It discloses that the result includes specific fields (summary, install snippet, etc.). While no extra behavioral traits (e.g., pagination, auth needs) are mentioned, the explanatory nature of the tool makes this acceptable.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single sentence that immediately conveys the main action and then lists the included content elements. No extraneous information, perfectly front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With one well-documented parameter and no output schema, the description sufficiently covers what the tool returns by enumerating the fields (summary, install snippet, use cases, FAQ, etc.). No gaps remain for effective use.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% and the slug parameter already has a detailed description (including example and source guidance). The tool description does not add further meaning about the parameter beyond what the schema provides, so baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves the full record for a single MCP server and enumerates its contents (summary, install snippet, etc.). It distinctively contrasts with siblings like search_mcp_servers (search interface) and compare_servers (comparison).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The slug parameter description suggests finding slugs via search_mcp_servers or list_top, providing a clear context for usage. However, the tool description itself does not explicitly state when to use this tool versus siblings, though the purpose strongly implies it's for detailed single-server info.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List the top-N MCP servers, ranked by composite quality score (default), raw GitHub stars, or real-world Smithery usage. Optional category filter.
| Name | Required | Description | Default |
|---|---|---|---|
| category | No | Optional category slug to filter. Same vocabulary as search_mcp_servers. | |
| limit | No | How many top servers to return (1-25). | |
| sort_by | No | Ranking signal: "score" (default, composite quality), "stars" (raw GitHub stars), "usage" (real Smithery uses in last 30 days). | score |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden. It discloses ranking methods and optional filtering but does not mention data freshness, authentication, rate limits, or edge cases like missing categories. This is adequate but not rich.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences front-load the action and ranking criteria, with no superfluous text. Every phrase earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's simplicity (3 optional params, no output schema), the description covers purpose, ranking options, and filter connection. It could briefly mention return format for completeness, but it is nearly adequate.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so parameters are well-documented. The description adds value by clarifying 'composite quality score' and linking category filter vocabulary to search_mcp_servers, extending meaning beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it lists top-N MCP servers, specifies ranking criteria (composite score, stars, usage), and mentions optional category filter. This distinguishes it from sibling tools like compare_servers, get_server_detail, and search_mcp_servers.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for ranking servers and links category filter vocabulary to search_mcp_servers, but it does not explicitly state when to use this tool versus alternatives or provide exclusion criteria.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search the mcpcatalogs.com directory by free-text query, optionally narrowing to a category. Returns ranked active MCP servers with summaries and links.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Free-text search query. Matches against server name, description, and AI-generated summary. Examples: "postgres", "browser automation", "notion read" | |
| category | No | Optional category slug to narrow results. Vocabulary: database, browser-automation, notion, slack, github, file-system, cloud-storage, search, ai-llm, communication, finance, productivity, developer-tools, monitoring, web-scraping, media, knowledge-graph, security, ecommerce, ops-infra, translation, other. | |
| limit | No | Maximum number of results to return (1-25). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries full disclosure burden. It states the tool searches and returns ranked active servers, implying read-only behavior. However, it does not explicitly mention safety, rate limits, or other behavioral traits.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences, front-loads the main action, and contains no extraneous information. Every sentence is meaningful.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has 3 parameters and no output schema, the description explains return format (ranked servers with summaries and links). It could mention result structure or pagination, but is adequate for a simple search tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% (baseline 3). The description adds value by providing examples for query, explaining matching fields, and listing the category vocabulary. This extra context helps the agent use parameters correctly.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool searches a directory by free-text query with optional category filtering, and returns ranked results. It distinguishes from sibling tools (compare_servers, get_server_detail, list_top) by focusing on search.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for searching, but does not provide explicit guidance on when to use this tool versus alternatives like list_top (for top servers) or get_server_detail. No exclusion criteria given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0compare_serversget_server_detaillist_topsearch_mcp_serversScored across 4 tools
Each tool serves a distinct purpose: comparing two servers, getting details of one, listing top servers, and searching. No overlap in functionality.
All tool names follow a consistent verb_noun pattern (compare_servers, get_server_detail, list_top, search_mcp_servers) with clear action verbs.
Four tools is well-scoped for a directory server, covering the main operations (search, detail, list, compare) without being excessive or insufficient.
The toolset covers core directory needs, but could benefit from a tool to browse categories or get server statistics, though the current set is largely complete.
Search and install 4,000+ security-scanned MCP servers from inside any MCP-aware AI client.
Capability registry for the agentic economy. Semantic search over verified MCP server listings.
Quality-ranked, cross-platform directory of AI coding skills, plugins and MCP servers.
Search a nightly-refreshed directory of MCP servers by keyword, category or topic.
