Loading ToolBench
MCP server for DeliverIQ email verification API — 12 tools for AI agents to interact with email verification, batch processing, email finding, domain intelligence (DNSBL, SPF/DKIM/DMARC, spam traps), and account management
Strong tool naming (verb_noun pattern consistently applied across all 12 tools). Descriptions are comprehensive (avg 180 chars, within 10-1024 baseline). All tools have documented input schemas with types and descriptions. However, output schemas are not explicitly documented in the source code provided, only inferred from descriptions. Parameter descriptions are detailed and include constraints (enums, defaults, ranges). No critical security issues detected (no secrets in params). Tool composition is sound: each tool has a single responsibility. Batch tools (deliveriq_batch_verify, deliveriq_batch_status, deliveriq_batch_download) form a coherent chain with proper ID threading. Error handling guidance is minimal, responses lack recovery hints.
Download batch verification results as CSV. Optionally filter by category (safe, risky, invalid, unknown). Returns a download URL valid for 24 hours.
Check the status of a batch verification job. Returns job status, progress, and summary statistics (safe, risky, invalid, unknown counts).
Submit a batch of 1-10,000 emails for verification. Returns a job ID for polling status. Optionally provide a webhook callback URL for completion notification.
Check a domain against 47 DNSBL zones to detect if it's listed on spam blacklists. Returns list of DNSBL hits with zone names and return codes.
Check current credit balance, usage breakdown, and plan information. Use this before performing credit-consuming operations to verify sufficient balance. Returns remaining credits, plan tier, billing period, and usage breakdown by category (single, batch, ESP sync). Free — no credit charge.
Get comprehensive domain intelligence including MX records, DNSBL status, infrastructure analysis, domain age, and trust score. Returns complete domain profile.
Output schemas not explicitly documented in source. Tool descriptions infer structure but LLMs cannot reliably parse undocumented responses. Batch operations return 'job ID for polling' but response structure is not formally specified.
Error handling lacks recovery guidance. Descriptions do not specify what errors are retryable, what user action is needed, or what to do if a lookup fails. E.g., deliveriq_find_email does not document behavior when no email is found.
deliveriq_batch_download category parameter uses enum but description does not explain what each category means or when to use them. LLM must infer from names alone.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 53 | 2026-07-28+ | v2 |
Find email addresses for a person at a company using first name, last name, and either domain or company name. Returns email address with confidence score and verification method.
Analyze a domain's email infrastructure: SPF, DKIM, DMARC, MTA-STS, BIMI, and TLS-RPT records. Returns configuration status and security assessment.
List all batch verification jobs with pagination and optional status filtering. Returns job summaries with status, email count, and completion percentage.
Get organization intelligence for a domain including company name, industry, employee count, and location. Free — no credit charge.
Evaluate an email address against 13 spam trap signals including honeypot patterns, role-based indicators, and disposable email detection. Returns risk classification and signal breakdown.
Verify a single email address with optional SMTP, Gravatar, HIBP, and deep intelligence checks. Returns reachability, syntax, MX, SMTP, and optional intelligence data.
deliveriq_find_email has mutually exclusive parameters (domain vs company_name) but description does not explicitly state this. LLM may pass both or neither, causing ambiguous failures.
No confirmation or dry-run pattern for write operations. deliveriq_batch_verify submits 1-10,000 emails for processing without a preview or confirmation step. Agents cannot validate intent before committing.