Can AI agents automate Dixa?

Dixa is automatable today through its REST API. This record is the Dixa customer-service platform (conversations, agents, queues, customers, tags, knowledge base, and analytics). It is not Zendesk, Gorgias, or Genesys Cloud. The v1 API server is https://dev.dixa.io. The OpenAPI description lists resources for analytics, agents, end users, conversations, tags, queues, webhooks, templates, business hours, knowledge articles, and the organization. Authentication is a bearer token in the Authorization header. Only administrators can create a token, under Settings, Integrations, API Tokens, and each token has selected permissions. The standards page sets a per-token limit of 10 requests per second, a burst of 4, and a quota of 864000 requests per day. Payloads are JSON. Outbound webhooks subscribe to conversation events. Duplicate delivery is possible, order is not guaranteed, 4xx responses are not retried, and 5xx responses are retried up to 3 times. Dixa also documents a hosted MCP server at https://dixa-mcp-public.fastmcp.app/mcp, authenticated with the same style of API key as Authorization Bearer. The getting started page lists 79 tools across analytics, conversations, agents, customers, tags, knowledge base, and settings, and documents Claude Desktop and Cursor setup. The same page marks the project beta, as-is, and not supported by Dixa's official support channels. The linked repository is https://github.com/Dixa-public/dixa-mcp-public. Pricing starts at €89 per agent. The pricing page says Dixa does not offer a self-serve trial. https://www.dixa.com/llms.txt is a first-party product index. Official language SDKs and a product CLI were not established. Connector catalogs were not opened. Computer-use viability is unassessed. API scores 7 and is the best path because the v1 catalog is the supported control plane. MCP scores 6 because it is a documented vendor preview that official support does not cover.

Best path today: api · Overall automatability: 7/10