MCP server exposing the full Gorgias helpdesk API to AI assistants
Server has 35 tools with consistent naming (verb_noun pattern: gorgias_*) and basic descriptions. However, parameter schemas are not visible in the provided source code, only tool names and descriptions are shown. Without access to actual Zod schema definitions or JSON Schema, schema quality cannot be verified. Error handling is present (safeHandler wrapper, sanitiseErrorForLLM), but descriptions are brief (10-50 chars) and lack actionable recovery guidance. No evidence of pagination parameters, output schema documentation, or per-parameter descriptions in the visible code. Access control is implemented (access-control.ts), which is a positive signal for security.
Add tags to a ticket
Create a message on a ticket (reply to customer or internal note)
Create a new ticket
List custom fields
List customers
List events
List integrations
Input schemas not visible in source code. Cannot verify parameter types, constraints, or descriptions. All 35 tools scored 0 for schema quality.
Tool descriptions are extremely brief (10-50 chars). Lack actionable context: when to use, what it returns, prerequisites, or recovery guidance. E.g., 'Create a new ticket' does not explain required fields, expected response, or error cases.
No evidence of pagination parameters (limit, offset, page_size, cursor) in list tools (gorgias_list_customers, gorgias_list_events, etc.). Large result sets will blow context window.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 37 | 2026-07-28+ | v2 |
List jobs
List macros
List rules
List satisfaction surveys
List tags
List teams
List messages on a ticket
List users
List views
List voice calls
List widgets
Remove tags from a ticket
Retrieve account information
Retrieve a customer by ID
Retrieve a file
Retrieve reporting statistics
Retrieve a ticket by ID
Search tickets
Set tags on a ticket (replace all existing tags)
Intelligent search across tickets with natural language understanding
Retrieve statistics with intelligent interpretation
Retrieve detailed ticket information with intelligent formatting
Update a custom field value on a customer
Update multiple custom field values on a customer
Update an existing message
Update an existing ticket
Update a custom field value on a ticket
Update multiple custom field values on a ticket
No output schema documentation visible. LLMs cannot plan downstream tool calls or extract required fields (e.g., ticket_id after create_ticket) without knowing response structure.
Error handling returns sanitized messages but lacks recovery guidance. 'An internal error occurred' tells LLM nothing about retry eligibility, user-fixable issues, or next steps.