buska-mcp
Buska MCP server: social listening for AI agents. Search public conversations on 30+ platforms, pull AI-scored buying signals, qualify leads. Remote, OAuth 2.1, 6 agent skills.
Documentation
Buska MCP Server
Social listening and lead generation for AI agents.
Give Claude, ChatGPT, Cursor or any MCP client the ability to search public conversations across 30+ platforms, pull AI-scored buying signals, and qualify leads.
·
·
·
·
What is Buska?
Buska is a social listening tool built for lead generation, not brand reports. It monitors Reddit, X (Twitter), LinkedIn, Hacker News, YouTube, Quora, Facebook, Instagram, TikTok, Bluesky, GitHub and 20+ more platforms in near real time, detects people who are actively asking for a product or service, and scores every mention from 0 to 100 for buying intent and fit with your ideal customer profile.
This repository documents the Buska MCP server, the Model Context Protocol endpoint that exposes all of that to AI agents. It is a remote, hosted server: nothing to install, nothing to run, sign in with OAuth 2.1 and your agent is connected.
| Endpoint | `https://api.buska.io/mcp` |
|---|---|
| Transport | Streamable HTTP |
| Auth | OAuth 2.1 (browser login, no API key to paste) |
| Tools | `search_mentions`, `get_signals`, `score_lead` |
| Works with | Claude Code, Claude Desktop, claude.ai, ChatGPT, Cursor, Windsurf, any MCP client |
| Plans | 7-day free trial (no card), then from $49/month |
Where to find it
- Smithery: smithery.ai/servers/buska-io/buska-mcp (one-click install for Claude, Cursor, VS Code and more)
- Cursor Directory: cursor.directory/plugins/buska-mcp (MCP server + the 6 skills, "Add to Cursor")
- Official MCP Registry: `io.buska/buska` (registry entry)
- Skills: `npx skills add Buska-io/buska-mcp` or browse buska.io/skills
Quick start
Claude Code (one line):
claude mcp add --transport http buska "https://api.buska.io/mcp"Cursor / Windsurf / Claude Desktop / ChatGPT: add a remote MCP server with the URL `https://api.buska.io/mcp` and sign in when prompted. Ready-made snippets are in `examples/` and a `.mcp.json` sits at the root of this repo for project-level configuration.
Then try:
> "Use Buska to find the 10 hottest buying signals for my keywords since yesterday, qualify each one against my ICP, and draft an opener for the top 3."
Tools
`search_mentions`
Live search of one platform for a keyword. Returns posts with content, author, URL and engagement metrics.
| Parameter | Type | Notes |
|---|---|---|
| `keyword` | string | 1 to 200 characters |
| `platform` | enum | `twitter`, `reddit`, `linkedin`, `youtube`, `hackernews`, `bluesky`, `stackoverflow`, `medium`, `substack`, `github`, `devto`, `mastodon`, `quora`, `pinterest`, `facebook`, `instagram`, `tiktok`, `producthunt`, `g2` |
| `limit` | number | 1 to 50, default 20 |
`get_signals`
The qualified buying signals Buska's engine already detected for your configured keywords, with the AI score and its reasoning.
| Parameter | Type | Notes |
|---|---|---|
| `keyword` | string | optional filter |
| `intent` | enum | `ACTIVE`, `COMPETITOR`, `PAIN`, `QUESTION`, `MENTION` |
| `platform` | string | optional filter |
| `minScore` | number | 1 to 10 |
| `since` | ISO date | only signals after this date |
| `limit` | number | 1 to 100, default 20 |
`score_lead`
Ask Buska to score any lead or post for buying intent and ICP fit.
Full reference with response shapes: buska.io/docs/mcp-server/tools. REST alternative: OpenAPI spec and API docs.
Agent skills (ready-made playbooks)
The `skills/` folder contains six playbooks in the `SKILL.md` format, so you can install them with the skills CLI:
npx skills add Buska-io/buska-mcp| Skill | What your agent does |
|---|---|
| find-leads-in-public-conversations | Daily shortlist of scored buying signals with a suggested opener per lead |
| daily-reply-shortlist | Picks the 5 conversations worth joining today and drafts a human reply for each |
| competitor-switch-alerts | Catches people complaining about a competitor or shopping for an alternative |
| qualify-leads-against-your-icp | Runs any list of leads through `score_lead` and sorts hot / nurture / ignore |
| surface-buying-objections | Mines real objections and buyer vocabulary to rewrite your landing page |
| fill-your-crm-from-social | Pulls, dedupes and qualifies signals, then routes them to your CRM |
Each skill is also published as a page with examples at buska.io/skills.
Why an MCP server for lead generation?
Most social listening tools count mentions. Buska finds buyers. Our own data shows why the two are different jobs: in a 60-day window of 451,903 public posts, only 1.4% carried strong buying intent, Reddit produced 77% of those hot posts, and LinkedIn had the highest buyer density of any social platform. The full study, with methodology and an open dataset, is at buska.io/research/state-of-buying-intent-2026.
An agent connected to Buska skips the noise: it asks for signals that are already scored, qualifies them against your ICP, and hands you the few conversations worth a reply.
Machine-readable resources
- `llms.txt`: https://www.buska.io/llms.txt
- `llms-full.txt` (full site content): https://www.buska.io/llms-full.txt
- Agent card: https://www.buska.io/.well-known/agent.json
- Citation guidance for AI assistants: https://www.buska.io/agent-permissions.json
- Registry manifest for this server: `server.json`
FAQ
Is the server free to use?
The MCP server is included with every Buska account, including the 7-day free trial (no credit card). Paid plans start at $49/month. See pricing.
Does Buska post or reply on my behalf?
No. Buska never publishes to social networks for you. Skills draft replies and openers; you review and send them.
Does it monitor private groups or DMs?
No. Buska only reads public posts.
Do I need to run anything locally?
No. This is a remote server. The repository only holds documentation, configuration examples and the skills.
Where do I report a problem?
Open an issue here or email tristan@buska.io.
License
The documentation, examples and skills in this repository are released under the MIT License. The Buska service itself is subject to the terms of service.
Frequently asked questions
What is buska-mcp?
buska-mcp is Buska MCP server: social listening for AI agents. Search public conversations on 30+ platforms, pull AI-scored buying signals, qualify leads. Remote, OAuth 2.1, 6 agent skills.
How do I install buska-mcp?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is buska-mcp open source?
Yes — it is hosted on GitHub at https://github.com/Buska-io/buska-mcp.
Related MCP tools
atlassian-mcp-serverOfficial remote MCP server for Atlassian. Securely connect Jira, Confluence, Jira Service Management, Bitbucket, and Compass to Claude, ChatGPT, Cursor, VS Code, and other AI tools using OAuth 2.1 or API tokens.
projectmemOpen-source coding agent memory. Records issues, attempts, fixes and decisions, then warns your agent before it repeats an approach that already failed. Native MCP server for Claude Code, Cursor, Antigravity and Codex. 100% local, no cloud, no telemetry. MIT.
Unity-MCPAI Skills, MCP Tools, and CLI for Unity Engine. Full AI develop and test loop. Use cli for quick setup. Efficient token usage, advanced tools. Any C# method may be turned into a tool by a single line. Works with Claude Code, Gemini, Copilot, Cursor and any other absolutely for free.
jcodemunch-mcpCut AI token costs 95%+ on code exploration. The leading MCP server for precise, symbol-level GitHub code retrieval via tree-sitter AST. Works with Claude Code, Cursor & any MCP client. 313B+ tokens saved.
meta-ads-mcpMeta Ads (Facebook/Instagram) MCP server for Claude, ChatGPT, Perplexity & Cursor — the Meta node of Pipeboard’s 5-platform family (+ Google, TikTok, Snap, Reddit). Hosted remote MCP, badged Meta Business Partner, free plan — no self-hosting required.
unity-mcpUnity MCP acts as a bridge between AI assistants and your Unity Editor. Give your LLM tools to manage assets, control scenes, edit scripts, and automate tasks within Unity.
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP