660+ data tools for agents: TikTok, Douyin, Instagram, YouTube, X, weather, GitHub. Pay per call.
MCPpedia last refreshed this data
io.github.satashili/toolsocket is an MCP server that 660+ data tools for agents: TikTok, Douyin, Instagram, YouTube, X, weather, GitHub. Pay per call. Its tool list has not been published yet over http and stdio, requires no API key, and scores 60/100 on MCPpedia's security, maintenance and efficiency rubric.
Config is the same across clients - only the file and path differ.
{
"mcpServers": {
"toolsocket": {
"env": {
"TOOLSOCKET_API_KEY": "ts_live_…"
},
"args": [
"-y",
"toolsocket",
"mcp"
],
"command": "npx"
}
}
}Are you the author?
Add this badge to your README to show your security score and help users find safe servers.
660+ data tools for agents: TikTok, Douyin, Instagram, YouTube, X, weather, GitHub. Pay per call.
This server supports HTTP transport. Be the first to test it - help the community know if it works.
Five weighted categories - click any category to see the underlying evidence.
No known CVEs.
Checked toolsocket against OSV.dev.
Be the first to review
Have you used this server?
Share your experience - it helps other developers decide.
Sign in to write a review.
Others in other
deja-vu: local memory over the session histories of forty-three coding agents.
MCP WordPress Remote proxy server
专门为 agent 打造的 agent 搜索工具,具备多语言搜索能力,覆盖中文/英文/学术/代码/购物/金融/新闻/百科。
Searchable football data provider documentation for AI coding agents. Like Context7 for football data.
MCP Security Weekly
Get CVE alerts and security updates for io.github.satashili/toolsocket and similar servers.
Start a conversation
Ask a question, share a tip, or report an issue.
Sign in to join the discussion.
660+ data tools for AI agents. One API key. Pay per call.
TikTok · Douyin · Instagram · YouTube · X · Reddit · Weibo · Bilibili · Zhihu · Kuaishou · Threads · LinkedIn · GitHub · Wikipedia · weather · FX · crypto …
ToolSocket is a remote MCP server plus a REST API. Your agent searches the catalog, inspects a tool's schema, and runs it. Every call is paid from one prepaid balance. You don't need an account with each platform, and there's no subscription.
This repository is the open-source CLI and stdio MCP bridge. The server lives at https://toolsocket.top.
The MCP endpoint is:
https://toolsocket.top/mcp
| Client | How |
|---|---|
| Claude.ai / Claude Desktop | Settings → Connectors → Add custom connector → paste the URL → sign in → Allow |
| ChatGPT | Settings → Apps & Connectors → Developer mode → Create → paste the URL, auth OAuth |
| Claude Code | claude mcp add --transport http toolsocket https://toolsocket.top/mcp, then /mcp → Authenticate |
| Cursor / VS Code / Windsurf / Codex | copy-paste configs at toolsocket.top/docs |
OAuth clients only need a sign-in. Clients that take headers can use an API key from the Console instead: Authorization: Bearer ts_live_….
npx -y toolsocket login # paste a ts_live_ key from toolsocket.top/console
npx -y toolsocket discover "douyin hot search"
npx -y toolsocket inspect openmeteo#forecast
npx -y toolsocket run openmeteo#forecast -i city=Tokyo -i days=2
npx -y toolsocket balance
Or install it globally with npm i -g toolsocket.
| Command | |
|---|---|
login [--key …] | Save your API key to ~/.config/toolsocket/config.json (mode 600) |
discover <query> [-n 10] | Plain-language search, in English or 中文 |
inspect <tool> | Description, input schema, example |
run <tool> [-i k=v]… | Run a tool. -i takes k=v, '{"json":true}' or @file.json |
get <run_id> | Fetch a run that is still in progress |
balance | Balance and today's spend |
mcp | stdio ⇄ remote MCP bridge (see below) |
logout | Forget the saved key |
Add --json to any command to get machine-readable output. Exit codes: 0 ok, 1 error, 2 the run failed, 3 balance too low.
Environment variables: TOOLSOCKET_API_KEY overrides the saved key. TOOLSOCKET_URL points the CLI at another server.
Some clients can only launch a local command (Cline, Zed, older Claude Desktop configs). For those, add:
{
"mcpServers": {
"toolsocket": {
"command": "npx",
"args": ["-y", "toolsocket", "mcp"],
"env": { "TOOLSOCKET_API_KEY": "ts_live_…" }
}
}
}
Your agent sees four meta-tools, not 660 tool definitions, so its context stays small:
| Tool | Does |
|---|---|
discover_tools(query, limit?) | Ranked search with price and live health |
inspect_tool(tool) | Full description, input JSON Schema, example |
run_tool(tool, input) | Runs the tool; waits up to 30 s, otherwise returns a run id |
get_run(run_id) | Status and output of a run |
The CLI has zero dependencies and is a single file: bin/toolsocket.js.
node bin/toolsocket.js --help
TOOLSOCKET_URL=http://127.0.0.1:8091 node bin/toolsocket.js discover weather
MIT