Loading ToolBench
A Model Context Protocol (MCP) server for name.com API integration, providing AI assistants with domain management capabilities
The name.com MCP server exhibits significant definition quality gaps across nearly all 19 tools. While tool names follow a clear verb_noun pattern (GetHelpResources, ManageDomains, etc.), the definitions lack the depth required for reliable LLM agent use. Most critically: (1) 13 of 19 tools (ManageDomains, ManageDNS, ManageDNSSECs, ManageEmailForwardings, ManageURLForwardings, ManageVanityNameservers, ManageTransfers, ManageOrders, ManageWebhookNotifications, ManageAccounts, ManageRefunds, and others) expose only a single 'operation' parameter with a string enum list in the description, no formal schema visible. This violates the constrained-input pattern and makes parameter validation impossible for agents. (2) No output schemas are documented anywhere in the source. Agents cannot predict what fields are returned, breaking downstream tool chaining. (3) Multi-operation tools like ManageDomains conflate 10+ distinct operations (list, create, get, update, purchaseprivacy, renew, setContacts, setNameservers, check, search) into a single tool, violating the single-responsibility principle and forcing agents to guess which operation to invoke and what parameters each accepts. (4) Tool descriptions are present but generic, they describe WHAT the tool does but not WHEN to use it, what prerequisites exist, or what structure is returned. (5) No error handling guidance is present; agents have no recovery path if an operation fails. The help tools (GetHelpResources, GetFeedbackLinks, GetTroubleshootingInfo) are well-named and have descriptions, but they are utility tools, not domain tools. The server reads from an OpenAPI spec (src/tool-generator.ts line references), but the flattened parameter approach sacrifices clarity and structure.
Returns the current account credit balance for the authenticated user.
Returns Feedback Links for rating the MCP server, feature requests, and bug reports on GitHub
Returns name.com Help Resources with links to documentation, support, and API reference
Returns Troubleshooting guidance including credential verification, domain ownership, account balance, and API rate limit information
Use the Hello endpoint to verify that your API connection and credentials are working properly. This simple call returns a success message (along with server time and version info) to confirm the API is reachable and authenticated.
Use Accounts endpoints (available upon request) to manage sub-accounts under your main account. For example, resellers can programmatically create new customer accounts with their own login credentials and permissions.
Multi-operation tools conflate 10+ distinct operations (e.g., ManageDomains supports list, create, get, update, purchaseprivacy, renew, setContacts, setNameservers, check, search). This violates the single-responsibility principle and forces agents to guess which operation to invoke, what parameters apply per operation, and what the result structure is for each operation.
No output schemas documented in source code. Agents cannot predict the structure of returned data, breaking tool chaining (e.g., if create_domain returns a domain_id, agents cannot reliably extract and pass it to subsequent calls). This violates the documented return types baseline (100% of A+ tools have documented return types).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 48 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 23 | - | v1 |
Multi-operation tool for DNS record management including list, create, get, update, and delete operations
Multi-operation tool for DNSSEC management including list, create, get, and delete operations
Multi-operation tool for domain information including TLD requirements and related operations
Multi-operation tool for domain management including list, create, get, update, purchaseprivacy, renew, setContacts, setNameservers, check, and search operations
Multi-operation tool for email forwarding management including list, create, get, update, and delete operations
Multi-operation tool for order management including list and get operations
Use the Refunds endpoint to delete eligible domains and advanced security products during the Add Grace Period (AGP) and automatically issue refunds for the associated order items.
Multi-operation tool for domain transfer management including list, create, get, and cancel operations
Multi-operation tool for URL forwarding management including create, list, get, update, and delete operations
Multi-operation tool for vanity nameserver management including list, create, get, update, and delete operations
Multi-operation tool for webhook notification subscription management including list, create, modify, and delete operations
APIs for working with Premium Domains.
Use TLD Pricing endpoints to retrieve general pricing information for your account.
Thirteen tools expose only a single 'operation' parameter with enum values listed in the description text, not in a formal JSON Schema enum constraint. Parameter validation is not machine-readable; agents must infer valid values from prose, inviting hallucinated or invalid operations.
Multi-operation tools do not document parameter dependencies. For example, ManageDomains 'create' operation requires different parameters than 'update', but agents see only the 'operation' string parameter with no description of which other parameters apply per operation. This violates the parameter-relationships check.
No error handling guidance documented. Tool descriptions do not explain what to do if an operation fails, whether errors are retryable, or what recovery steps are available. This violates the recovery-guide and error-classification patterns.
Tool descriptions are generic and do not explain WHEN to use a tool vs. similar alternatives, what prerequisites exist (e.g., 'requires authenticated API key'), or what domain-specific guidance applies. This falls short of the 'prompt-engineering' description baseline (WHAT, WHEN, prerequisites).
ManageRefunds is marked DESTRUCTIVE (deletes domains, issues refunds) but has no confirmation-request pattern or dry-run support documented. Agents can trigger irreversible refunds without safeguards.
No pagination documented for tools that likely return lists (ManageDomains list operation, ManageOrders list operation, etc.). Tool descriptions do not mention limit, offset, or next_cursor parameters. This violates the paginated-result pattern and risks context window exhaustion.