Screen candidates
screen_domain_candidatesBatch-check registry status and use standard TLD costs for an initial budget screen. viable does not mean a registrar can complete the order.
MCP · REST · OpenAPI 3.1
Domain evidence for agent workflows.
NamePico provides three read-only MCP tools that split candidate screening, live price verification, and deep evidence research into controlled steps. No API key is required, and unknown, timeout, and partial-coverage states remain explicit.
REMOTE MCP ENDPOINTREAD-ONLY
https://namepico.com/mcpNo API key
Stateless
Schema 1.3
The MCP service identifier is namepico; legacy entry points, tool names, and existing fields remain compatible.
Start with the MCP endpoint.
Add this site's /mcp URL as a remote MCP server. It uses Streamable HTTP, keeps no session state, and handles initialization, tool listing, and calls at the same endpoint.
ENDPOINT
https://namepico.com/mcp
Supports GET, POST, DELETE, and CORS. No Authorization header is required.
RECOMMENDED FLOW
Generate 10–20 candidates, screen them, verify prices, then research only 1–3 finalists. Do not run expensive reports for the entire batch.
Three tools, one explicit call sequence.
Every tool is declared read-only but accesses public internet sources. A single source failure is not disguised as a successful batch result.
Verify registrar prices
verify_domain_pricesVerify domain-specific quotes, premiums, and suspected reserved states only for shortlisted candidates, with source, timestamp, and cache state.
Research finalists
get_domain_reportInvestigate registration, website, history, security, brand signals, and similar domains. Formal trademark and legal conclusions remain out of scope.
Facts, judgments, and unknowns stay separate.
Prefer canonical fields when reading responses. Legacy clients can continue reading existing fields and ignore additions.
| Layer | Canonical field | Meaning |
|---|---|---|
| Registry status | registryStatus | Public registration state: available / registered / reserved / unknown |
| Orderability | registrarOrderability | Whether a registrar returned a domain-specific quote, separate from RDAP state |
| Risk | verdict.score | Rates only the evidence obtained; it is not trademark or security certification |
| Coverage | riskCoverage | Completed checks, failed sources, and unverified items |
| Brand | brandEvidence | Observed evidence with source and timestamp, deduplicated by evidenceId |
| Conclusion | brandEvidenceStatus.conclusion | Remains not_determined; no legal conclusion is fabricated |
Important: exact_string only means that a field contained the same string. Also read
matchedField, isStandaloneEntity and isBrandNameMatch. A source returning no_match does not prove that no conflict exists.Use the same evidence without MCP.
REST, plain text, and web reports share the same fact layer. These URLs come from the current deployment environment and do not assume an unreleased domain.
SINGLE-DOMAIN JSON
curl https://namepico.com/report/namepico.com.jsonBATCH REGISTRY SCREEN
curl -X POST https://namepico.com/api/v1/domains/check \
-H 'content-type: application/json' \
-d '{"names":["namepico","namelens"],"tlds":["com","dev"]}'Boundaries by design.
Limits protect free upstream sources and prevent candidate research from becoming a reconstructable query history.
8 per minute
Single-domain queries are limited per source; up to 30 detail reports per hour.
300 per hour
Batch quota is counted by domain, not by HTTP request.
Candidates are not stored
Candidate names are not written to the application database; only aggregate operating metrics are retained.
When the API returns 429, read retry-after. When status is unknown, pricing is unverified, or coverage is partial, the caller must preserve that uncertainty.