Dollie is an AI operations platform for WordPress agencies and managed service providers. Its core idea is to first structure "agency context," then let AI perform operations in real business scenarios.
I. Purpose Dollie is designed for teams that need to manage multiple client WordPress sites simultaneously. It consolidates client profiles, site information, workflows, and standards into a unified private cloud environment, then provides this context to AI tools such as Claude, Cursor, and VS Code through an MCP server. The goal is to eliminate the need for AI to re-learn who the client is and what the tech stack is every time, and instead have it work directly based on existing client history, site status, and agency standards.
II. Main Capabilities
- Client management: Add client profiles, preferences, and project history, and define the agency's own ways of working and standards; supports importing clients from Stripe or CSV; AI can answer client-related questions based on real data, such as which sites under a given client need updates, or draft monthly reports in the client's preferred tone.
- Site connection and monitoring: Supports connecting WordPress sites on any host, and viewing updates, security, availability, and errors across the entire site portfolio in a single interface. Operations such as site health checks, uptime monitoring, plugin update history queries, security scans, and error log reading can all be triggered through conversation.
- Context interoperability: Brings client, site, and operational context into Claude, Cursor, VS Code, or Dollie's own apps through an MCP server (30-second connection, 30+ tools); also provides an OpenAI-compatible API and SDK for building custom chat interfaces, client portals, or internal dashboards, and supports connecting to external MCP services such as GitHub, Slack, and Firecrawl.
- AI workspace: Launch a complete WordPress environment in about 30 seconds, automatically loading client context, agency standards, and deployment targets; supports reusing configurations with blueprint templates; agents can read files, write code, install plugins, test themes, and create prototype demos within the workspace, with WP-CLI, SSH, and database access, and all operations are audited and can be rolled back.
- Real site operations: Dollie's agents do more than provide suggestions—they directly perform plugin updates, run security scans, read error logs, execute WP-CLI commands, and more; supports working with a variety of models including Claude, GPT-4, Gemini, and others.
III. How to Use First, add clients, sites, and processes in Dollie to establish agency context; then bring that context into the AI tools your team already uses through the MCP endpoint, or use Dollie's chat and desktop apps directly; issue tasks in the conversation, such as checking a site's health and availability, pulling plugin update history, or drafting a personalized report and sending it by email. All team members share the same context, and client and site information stays consistent across different tools.
IV. Use Cases Suitable for agencies and managed teams that need to manage WordPress sites in bulk, generate client reports, perform site inspections and maintenance, and quickly set up demo environments for clients; also suitable for teams that want to build custom portals or internal tools based on their own client data.
V. Limitations and Not Yet Available The official page states that the complete feature overview is still under construction. Details regarding billing, pricing, the specific list of supported hosting providers, and the availability of individual features are not yet provided and are subject to official information released later.
