Crawdar MCP server

• Registry namecom.crawdar/business-researchverified in the MCP registry

The registry name sits under crawdar.com. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

• Stateanswerschecked 2026-10-24
• Tools10
• Server calls itselfcrawdar
• Server version1.5.0
• Protocol version2025-06-18

On 2026-10-24 the server answered initialize and tools/list without a sign-in. First answer in 59 ms.

• Its tools do every use case we checked: find businesses whose customers repeatedly report the same problem; read review evidence behind recurring complaints before pitching.
• Checked on 3 of the 7 days it takes to judge reliability.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

• INITE MCP scoreafter a weeka score needs 7 daily checks: 3 so far
• What an agent can do with it2 of 2 use cases
• Answeredmeasuring3 of the 7 daily checks it takes to print a share
• Median first answer215 ms
• Space in an agent's context≈ 2,264 tokensthe definitions of all its tools (10), which an agent carries on every turn
• Tool definitionsA · 98 of 100
• Typed results10 of 10 toolsdeclared outputSchema: results come typed, read once

• Parameters without a description: in 1 of 10

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (30%), doing the product's job (25%), definitions (20%), context (10%) and use (15%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

• scan_review_pain_points

  Scan a named business or discover up to three businesses for recent recurring complaints. Returns individually sourced reviews and possible technical solutions. Coverage is sampled; unavailable reviews never mean no problems. Five scans…

• research_businesses

  Use for a quick synchronous search when the user supplies one natural-language brief and can wait for results in the current tool call. For durable progress, retry, export, or large agent workflows, use start_lead_search instead.

• search_businesses

  Use for a quick synchronous search when target, geography, and qualification fields are already structured. Use research_businesses for a plain-language brief, or start_lead_search for durable asynchronous work.

• sandbox_businesses

  Return deterministic fictional businesses without consuming provider capacity. Use this to test schemas, clients, and agent loops.

• start_lead_search

  Use for the recommended durable asynchronous workflow. Creates a private queued search job and returns a job id plus jobToken for progress polling, cursor paging, refinement, retry, cancellation, and CSV or JSON export.

• get_lead_search

  Read job status or one cursor-paginated result page. Keep jobToken private and pass nextCursor back unchanged.

• refine_lead_search

  Create a new private search job by changing the brief or criteria of a completed job. The original job stays unchanged.

• retry_lead_search

  Create a replacement job when a prior job failed or exceeded its execution window. The original job stays available for audit.

• cancel_lead_search

  Cancel a queued or running search job. Completed and failed jobs remain immutable.

• explain_crawdar

  Get result semantics, limits, interface URLs, and safe operating guidance. This tool does not consume a search.