MCP Available Tools

The Buska MCP server exposes 3 tools that AI assistants can call:

1. search_mentions

Search social mentions by keyword and platform. Equivalent to the POST /v1/mentions/search API endpoint.

ParameterTypeRequiredDescription
keywordstringYesSearch term
platformstringNoPlatform filter
limitnumberNoMax results (default 20)

2. score_lead

AI-powered B2B lead qualification. Equivalent to the POST /v1/leads/score API endpoint.

ParameterTypeRequiredDescription
contentstringYesText to score (min 10 chars)
keywordstringNoKeyword for context
contextstringNoProduct/service description

3. get_signals

Retrieve qualified leads with filters. Equivalent to the GET /v1/signals API endpoint.

ParameterTypeRequiredDescription
keywordstringNoFilter by keyword
intentstringNoFilter by intent type
platformstringNoFilter by platform
minScorenumberNoMinimum score (0-10)
sincestringNoISO date filter
limitnumberNoMax results (default 20)

Example usage with Claude

Once the MCP server is connected, you can ask Claude natural language questions like:

  • "Search Twitter for people looking for a CRM tool and score the results"
  • "Find the top 5 hot leads from LinkedIn this week"
  • "Score this Reddit post for buying intent: [paste text]"
  • "Get all ACTIVE signals with a score above 7 from the last 24 hours"

Practical example: multi-step AI agent workflow

Try asking Claude a complex query that chains multiple tools together:

text
Ask Claude: "Search Twitter for people looking for a CRM tool, score the top results, and show me the ones with buying intent above 70"

Claude will automatically:

  1. Call search_mentions with keyword "CRM tool" on Twitter
  2. Call score_lead for each result to get detailed scoring
  3. Filter results where the score exceeds 70
  4. Present a summary table with the qualified leads, their scores, and reasoning
The MCP server uses the same API key and rate limits as the REST API. Each tool call counts as one API request toward your monthly quota.

Related guides