CLI tool and MCP server for ExportComments.com API - export comments and reviews from 40+ platforms. Optimized for AI agent usage.
Strong tool naming (verb_noun pattern), comprehensive parameter schemas with Zod validation, and detailed descriptions. All 7 tools have clear names and descriptions (avg 150+ chars). Input schemas are well-typed with optional/required distinction. However, output schemas are not documented, responses are JSON-stringified without field-level documentation. Error handling returns JSON but lacks actionable recovery guidance. Missing tool annotations (readOnlyHint/destructiveHint). Parameter descriptions are good but some lack format constraints (e.g., date formats, VPN country codes).
Check the status of an export job by its GUID. Returns full job details including status, progress, download URLs, and error info. Job statuses: queueing → progress → done | error. Set wait=true to poll until the job reaches a terminal state.
Detect which platform a URL belongs to and return supported options. Use this before export_comments to understand what options are available.
Download a completed export's data for analysis (status must be "done"). Returns JSON for Premium/Business accounts and CSV for Free/Personal accounts — CSV is plain text, ideal for reading and analyzing the comments directly. The format defaults to the best option for your plan; pass "format" to override (JSON requires a Premium or Business plan).
Create a new export job to extract comments or reviews from a URL. Supports 33+ platforms: Instagram, YouTube, TikTok, Facebook, Twitter/X, Reddit, LinkedIn, Trustpilot, Amazon, Yelp, Google Reviews, and more. Returns a job GUID for tracking. Use check_export to monitor progress. Set wait=true to poll until completion (up to 10 minutes).
Get the authenticated user's profile information including plan type, API access, and usage statistics.
Output schemas not documented. Tools return JSON.stringify(result) without field-level documentation. LLMs cannot plan downstream calls or extract specific fields reliably.
Error responses lack actionable recovery guidance. Errors return raw JSON without suggesting next steps (e.g., 'Try detect_platform() first' or 'Check your API token').
Tool annotations missing. No readOnlyHint, destructiveHint, or idempotentHint declared. LLMs cannot infer which tools are safe to retry or have side effects.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 69 | 2026-07-28+ | v2 |
List all export jobs for the authenticated account with pagination.
List all 33+ supported platforms with their URL patterns, export options, and example URLs.
Parameter descriptions lack format constraints. 'vpn' accepts country names but no enum or pattern specified. 'min_date'/'max_date' mention ISO 8601 but no validation example. 'limit' unbounded (no min/max).
list_exports and list_platforms descriptions are minimal (60-70 chars). 'List all export jobs' and 'List all 33+ supported platforms' lack context on when to call them or what structure they reveal.