AgentScout
Read-only search over 1,706 AI agent tools and 239 MCP servers. No sign-up needed.
我該用這個嗎
品質與安全性
根據工具定義與協定合規性的自動化分析。
上下文成本
這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高，可用於其他工作的注意力就越少。
安裝
一鍵安裝
將以下內容加入你的 `claude_desktop_config.json` 檔案：
{
"mcpServers": {
"directory": {
"url": "https://agentscout-mcp.agentscouthq.workers.dev/mcp"
}
}
}遠端端點
https://agentscout-mcp.agentscouthq.workers.dev/mcpstreamable-http它能做什麼
工具清單
工具（4）
🟢search_tools(query, category, pricing, open_source, runs_locally, ...)
Search the AgentScout directory of AI agent tools and MCP servers by keyword, with optional category, pricing, open-source and data-locality (runs_locally) filters. Each result has data_locality (offline, self_hosted, cloud or unknown) with a plain note. Returns up to 25 matches with pricing as the maker states it and the AgentScout listing URL.
輸入結構描述
{
"type": "object",
"properties": {
"query": {
"type": "string",
"maxLength": 200,
"description": "Keywords, e.g. 'browser automation' or 'postgres'. Optional if a filter is set."
},
"category": {
"type": "string",
"maxLength": 80,
"description": "Exact category name from list_categories, e.g. 'MCP servers'."
},
"pricing": {
"type": "string",
"enum": [
"any",
"free",
"freemium",
"paid",
"unknown"
],
"description": "Pricing type. 'freemium' = free plan plus paid plans."
},
"open_source": {
"type": "boolean",
"description": "true = only open-source tools."
},
"runs_locally": {
"type": "string",
"enum": [
"offline",
"self_hosted_or_offline",
"any"
],
"description": "Where the tool can run, so data can stay on your side. 'offline' = runs fully on your own machine (e.g. with local models). 'self_hosted_or_offline' = also includes tools you can deploy on your own server. 'any' (default) = no filter. Based on what each listing says; tools without clear evidence are excluded from both filters."
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 25,
"description": "Max results (default 10)."
}
},
"additionalProperties": false
}🟢get_tool(slug)
Get the full AgentScout listing for one tool by its slug (from search_tools results), including key features, platforms, GitHub repo and stars, and data_locality with the evidence phrase.
輸入結構描述
{
"type": "object",
"properties": {
"slug": {
"type": "string",
"pattern": "^[a-z0-9][a-z0-9-]{0,99}$",
"description": "Tool slug, e.g. 'cursor'."
}
},
"required": [
"slug"
],
"additionalProperties": false
}🟢list_categories
List every AgentScout category with its tool count and page URL.
輸入結構描述
{
"type": "object",
"properties": {},
"additionalProperties": false
}🟢best_for_task(task, pricing, open_source, runs_locally, limit)
Suggest AgentScout tools for a task described in plain words (e.g. 'scrape websites with an agent'), ranked by relevance with listed GitHub stars as a tiebreak. Optional pricing, open-source and runs_locally filters; use runs_locally when data must stay on the user's machine or servers (e.g. 'no-code automation that runs locally').
輸入結構描述
{
"type": "object",
"properties": {
"task": {
"type": "string",
"minLength": 2,
"maxLength": 200,
"description": "What you want to get done."
},
"pricing": {
"type": "string",
"enum": [
"any",
"free",
"freemium",
"paid",
"unknown"
]
},
"open_source": {
"type": "boolean"
},
"runs_locally": {
"type": "string",
"enum": [
"offline",
"self_hosted_or_offline",
"any"
],
"description": "Where the tool can run, so data can stay on your side. 'offline' = runs fully on your own machine (e.g. with local models). 'self_hosted_or_offline' = also includes tools you can deploy on your own server. 'any' (default) = no filter. Based on what each listing says; tools without clear evidence are excluded from both filters."
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 10,
"description": "Max results (default 5)."
}
},
"required": [
"task"
],
"additionalProperties": false
}社群
證據