Start building a new agent immediately with a pre-provisioned Salesforce Org — no Agentforce or Data 360 configuration needed.

Connect your Salesforce Org and build on Agentforce in the tool that best fits your needs.

Connect your AI agent to the platforms your customers use most. Deploy your agent across web, SMS, and social APIs like WhatsApp and Messenger.

Automate agent development with Agentforce Arc, an experiment in "agents building agents." ARC generates complete build plans—including sub-agent orchestration, logic, guardrails, and smoke tests—directly within your Salesforce environment.

A suite of open-source skills spanning the full agent development lifecycle — from authoring and testing to observability.

An open-source, unified DSL for defining Agentforce agents. Use a single .agent file to orchestrate instructions, subagents, and deterministic rules, giving you full control over how LLMs execute business logic.

Equip Claude Code to manage Agentforce Grid workbooks directly. This experiment combines specialized AI skills for grid configuration with an MCP server that provides the tools for real-time CRUD operations.

Autonomous AI agent runtime with goal-driven workflows, 28 built-in tools, and a REST API — built natively on Salesforce.

Give your Agentforce agents the ability to reference documents when answering questions. No vector databases or complex setup required.

The next generation of Robotic Process Automation (RPA). GUI Process Automation (GPA) is an agent that learns tasks (submitting receipts, logging notes, etc.) by watching a user perform them once.

A post-processing layer that wraps MCP clients and filters tool outputs down to only the relevant context, cutting inference costs without sacrificing reasoning performance.

A skill-centric orchestration framework that routes tasks to the right agent based on competence and cost — no massive routers needed.

Browse Salesforce documentation directly from your IDE using the Model Context Protocol.