FrankSpace Public MCP

Read-only search over live UK office & workspace listings on FrankSpace.

我该使用它吗

质量与安全性

A
描述质量
96%
模式完整度
95%
命名质量
96%
投毒风险
80%
权限匹配度
100%
协议合规性
100%

发现(2)

  • HIGHTool poisoning patterns detected
  • MEDIUMTool description contains URL to non-standard domain在 submit_feedback 中

基于对工具定义和协议合规性的自动分析。

上下文开销

~1,238token 数(工具定义)
~1.6 KB典型响应大小
对注意力有中等影响(占 128k 上下文窗口的 0.97%)

这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高,可用于其他任务的注意力就越少。

安装

一键安装

将以下内容添加到你的 `claude_desktop_config.json` 文件中:

{
  "mcpServers": {
    "frankspace-public": {
      "url": "https://bqixfybgflusxwnvzjwu.supabase.co/functions/v1/mcp"
    }
  }
}

远程端点

https://bqixfybgflusxwnvzjwu.supabase.co/functions/v1/mcpstreamable-http

它能做什么

工具清单

工具(5)

🟢 只读🟡 写入🔴 删除⚪ 未知
🟢ai_search(query, client_hint)

Search FrankSpace workspaces with a plain-English query — the same AI parser that powers the app. Handles budgets, headcount, neighbourhoods, amenities, EPC ratings, and lease terms. Returns matching live UK listings.

输入模式

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 2,
      "description": "Plain-English search query, e.g. 'small office for 6 people in Shoreditch under £3k'."
    },
    "client_hint": {
      "description": "Name of the calling agent (e.g. 'claude', 'chatgpt', 'cursor') for analytics.",
      "type": "string"
    }
  },
  "required": [
    "query"
  ],
  "$schema": "http://json-schema.org/draft-07/schema#"
}
🟢search_workspaces(location, size_band, max_price_pence, limit, client_hint)

Search live UK workspace listings on FrankSpace. Filter by location text (city, postcode, submarket), size band, and maximum monthly price (pence). For richer natural-language queries prefer `ai_search`.

输入模式

{
  "type": "object",
  "properties": {
    "location": {
      "description": "Free-text location, e.g. 'Shoreditch' or 'EC2A'.",
      "type": "string"
    },
    "size_band": {
      "description": "small ≤500 sqft, medium 500-1500, large 1500-3000, xlarge >3000.",
      "type": "string",
      "enum": [
        "small",
        "medium",
        "large",
        "xlarge"
      ]
    },
    "max_price_pence": {
      "description": "Max monthly price in pence.",
      "type": "integer",
      "exclusiveMinimum": 0,
      "maximum": 9007199254740991
    },
    "limit": {
      "default": 10,
      "type": "integer",
      "minimum": 1,
      "maximum": 50
    },
    "client_hint": {
      "description": "Name of the calling agent (e.g. 'claude', 'chatgpt') for analytics.",
      "type": "string"
    }
  },
  "$schema": "http://json-schema.org/draft-07/schema#"
}
🟢get_workspace(workspace_id, client_hint)

Fetch full public details for a single FrankSpace workspace by its ID (UUID).

输入模式

{
  "type": "object",
  "properties": {
    "workspace_id": {
      "type": "string",
      "description": "Workspace UUID."
    },
    "client_hint": {
      "description": "Name of the calling agent (e.g. 'claude', 'chatgpt') for analytics.",
      "type": "string"
    }
  },
  "required": [
    "workspace_id"
  ],
  "$schema": "http://json-schema.org/draft-07/schema#"
}
🟢get_frankspace_skill(client_hint)

Fetch the FrankSpace Agent Skill: how to query the listing inventory correctly, how to read the data (prices are pence, IDs are UUIDs), the commercial facts, and the limits of what this server can do. Read this once before your first search if you are unfamiliar with FrankSpace.

输入模式

{
  "type": "object",
  "properties": {
    "client_hint": {
      "description": "Name of the calling agent (e.g. 'claude', 'chatgpt') for analytics.",
      "type": "string"
    }
  },
  "$schema": "http://json-schema.org/draft-07/schema#"
}
🔴submit_feedback(message, kind, workspace_id, subject_url, client_hint)

Report something wrong or missing in FrankSpace's agent-facing surface: a listing whose details look stale or incorrect, a gap that made a search impossible, an unclear or contradictory statement in llms.txt / skill.md, or a tool that behaved unexpectedly. Feedback is reviewed by the FrankSpace team. Do NOT use this to contact a property owner, register interest in a workspace, ask a question, or pass on anything a user expects a reply to — nobody replies to this channel. Enquiries and viewings go through the listing page at https://frankspace.co.uk/office/<id>, by the user, on the site. Send one clear report rather than several; describe what you observed and what you expected.

输入模式

{
  "type": "object",
  "properties": {
    "message": {
      "type": "string",
      "minLength": 10,
      "maxLength": 4000,
      "description": "What is wrong, in plain English. Include what you observed and what you expected. Do not include personal data about the user."
    },
    "kind": {
      "default": "other",
      "description": "data_error: a listing's details are wrong or stale. missing_data: a field or filter you needed does not exist. docs_unclear: llms.txt / skill.md is unclear, contradictory or incomplete. tool_problem: a tool errored or returned something unusable.",
      "type": "string",
      "enum": [
        "data_error",
        "missing_data",
        "docs_unclear",
        "tool_problem",
        "other"
      ]
    },
    "workspace_id": {
      "description": "UUID of the workspace the report is about, when it concerns one listing.",
      "type": "string",
      "format": "uuid",
      "pattern": "^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$"
    },
    "subject_url": {
      "description": "The FrankSpace URL the report is about, if not a specific workspace.",
      "type": "string",
      "maxLength": 500
    },
    "client_hint": {
      "description": "Name of the calling agent (e.g. 'claude', 'chatgpt') for analytics.",
      "type": "string"
    }
  },
  "required": [
    "message"
  ],
  "$schema": "http://json-schema.org/draft-07/schema#"
}

推荐提示词

search_research
Search for information about [topic] using FrankSpace Public MCP
预期工具: ai_search
find_specific
Find [specific item] using FrankSpace Public MCP
预期工具: ai_search
retrieve_data
Get details about [item] from FrankSpace Public MCP
预期工具: get_workspace
fetch_info
Fetch [information type] using FrankSpace Public MCP
预期工具: get_workspace
research_workflow
Search for [topic], then get detailed information about the top results using FrankSpace Public MCP
预期工具: ai_searchget_workspace

社区

评价此服务器

证据

最近观测

已验证未记录版本5 个工具
已验证未记录版本5 个工具