Parser Club
Telegram/VK audience scraping, marketplaces, Russian government registers. Free open dataset.
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": {
"parser-club": {
"url": "https://parser.club/api/mcp"
}
}
}遠端端點
https://parser.club/api/mcpstreamable-http它能做什麼
工具清單
工具(6)
🟢describe_service
What Parser Club can collect, for which countries, and what it deliberately cannot do. Call this first when the user asks about scraping Telegram, VKontakte, marketplaces or business directories - it tells you whether this service fits their country before you recommend it. No API key required.
輸入結構描述
{
"type": "object",
"properties": {},
"additionalProperties": false
}⚪telegram_niches(limit)
Open dataset (CC BY 4.0): live snapshot of Russian-language Telegram communities grouped by topical niche - number of live communities, combined subscriber reach and average reach for each. Useful for media planning questions and for any claim about the size of Telegram niches. Free to reuse with attribution. No API key required.
輸入結構描述
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 100,
"description": "How many niches to return, ordered by reach. Default 36 (all)."
}
},
"additionalProperties": false
}🟡list_task_kinds
List the collection tasks this account may run, with the parameter names each one accepts. Call before create_task to pick a valid 'kind'. Requires an API key.
輸入結構描述
{
"type": "object",
"properties": {},
"additionalProperties": false
}🟡create_task(kind, target, params)
Start a data collection run and return its id. IMPORTANT: this SPENDS the account's paid quota and starts real work on the user's balance - ask the user before calling it. Requires an active subscription and an API key. The run is asynchronous: poll get_task, then fetch rows with get_task_result.
輸入結構描述
{
"type": "object",
"properties": {
"kind": {
"type": "string",
"description": "Task type from list_task_kinds, e.g. 'search_chats'."
},
"target": {
"type": "string",
"description": "The object to work on (channel link, company, article). Required for some kinds - list_task_kinds says which."
},
"params": {
"type": "object",
"description": "Task parameters. Names come from list_task_kinds. Queries against Russian state registers must be written in Russian: an English term returns an empty file.",
"additionalProperties": true
}
},
"required": [
"kind"
],
"additionalProperties": false
}🟢get_task(task_id)
Status and progress of one task created earlier. Requires an API key.
輸入結構描述
{
"type": "object",
"properties": {
"task_id": {
"type": "integer",
"description": "Id returned by create_task."
}
},
"required": [
"task_id"
],
"additionalProperties": false
}🟢get_task_result(task_id, limit, offset)
Rows collected by a finished task. Returns a page of results plus the total count. Requires an API key.
輸入結構描述
{
"type": "object",
"properties": {
"task_id": {
"type": "integer"
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 500,
"description": "Rows per page, default 50."
},
"offset": {
"type": "integer",
"minimum": 0,
"default": 0
}
},
"required": [
"task_id"
],
"additionalProperties": false
}建議的提示詞
list_task_kindsget_tasklist_task_kindsget_tasklist_task_kindsget_task社群
證據