{"name":"AgentCrush MCP Server","version":"1.0.0","api_version":"v1","protocol":"MCP","protocolVersion":"2024-11-05","description":"Read-only AI agent market intelligence. Evidence-ranked across 4 category indices: model families, tokenized agents, service agents, developer agents. Live deployment, citation, on-chain, and adoption signals.","tools":[{"name":"search_agents","description":"Search AI agents by name or keyword. Returns matching agents with category, tier, and rank info. Use the `filters` object for structured constraints; future versions will add more filter keys without breaking the API."},{"name":"get_agent_details","description":"Get full details for an AI agent including all category scores it qualifies for (model_family, tokenized, service, developer). Returns identity, raw signals, sub-scores, evidence-ready status across all relevant categories."},{"name":"get_agent_history","description":"Get rank and score history for an AI agent over time. Returns daily snapshots up to 90 days. Useful for showing how an agent's standing has evolved."},{"name":"compare_agents","description":"Compare 2-5 agents side-by-side across all their categories. Returns scoring data per agent and an automatic comparison summary."},{"name":"list_categories","description":"List the 4 AgentCrush agent categories with counts, methodology versions, and evidence-ranked agent counts. Use this to discover what kinds of agents AgentCrush tracks."},{"name":"get_category_ranking","description":"Get the full ranking for a specific category. Returns agents ordered by composite score with all sub-scores visible."},{"name":"get_methodology","description":"Get the scoring methodology for a category — weights, signal sources, formulas, evidence-ready rule, and known limitations. Use this when explaining HOW a ranking works to an end user. Methodology travels with data so LLMs can answer \"how is this ranked\" accurately."}],"endpoint":"POST https://www.agentcrush.xyz/api/mcp/v1","docs":"https://www.agentcrush.xyz/developers/mcp","rate_limit":"60 requests/minute per IP"}