WP Agents – AI Agent Framework for WordPress

Build autonomous, hook-driven agents for WordPress — automate tasks and add LLM intelligence with clean, developer-first architecture.

📣 The project is open to contributions. Developers are invited to test, improve, and expand the framework as it evolves.

Key Features

• Providers: Define and manage LLM providers (e.g., OpenAI, Anthropic) through a unified API for completions and structured outputs.
• Tools: Register callable functions that agents can dynamically execute during reasoning or task completion.
• REST API: Provides endpoints that allow external applications to interact with agents and send or receive chat messages programmatically.
• Memory: Supports persistent conversation history, enabling agents to retain context across sessions and handle long-running or multi-turn interactions.
• Agents: Create modular agents that interact with WordPress using a consistent architecture.
• Tests: Basic Pest testing structure in place — full coverage and assertions to be implemented.

Quick Start

cd wp-content/plugins

# Clone the plugin into your plugins folder
git clone git@github.com:santerref/wp-agents.git wp-agents

# Install PHP packages
cd wp-agents && composer install

# Install the plugin with WP-CLI or via the administration
wp plugin activate wp-agents

REST API

Each agent can be accessed programmatically through the WordPress REST API.

This allows external systems or front-end applications to send a message to any registered agent and receive the model’s response in JSON format.

Endpoint

POST /wp-json/wp-agents/v1/chat

Example

curl --location 'https://example.com/wp-json/wp-agents/v1/chat' \
--header 'Content-Type: application/json' \
--data '{
    "agent": "taxonomy_agent",
    "message": "Hello"
}'

Example response

{
    "agent": "taxonomy_agent",
    "message": "Hello",
    "response": {
        "role": "assistant",
        "content": "Hello! How can I assist you today?"
    }
}

Demo

We now provide an admin UI that lists all registered agents. From there, you can activate the built-in Post Taxonomy Agent to see a working example in action.

Roadmap

• Workflows: Introduce workflows to connect and orchestrate multiple agents for complex, multi-step tasks.
• Tests: Expand and complete Pest test coverage for agents, tools, and provider logic.
• Providers: Add more LLM providers and improve compatibility with third-party APIs.
• RAG: Integrate Retrieval-Augmented Generation to allow agents to query custom datasets, documents, or WordPress content for more context-aware responses.