Headless Toolkit
Bring Salesforce to every agent, app, and experience.
Salesforce becomes available wherever users build and work, grounded in your business logic and governed everywhere.
Salesforce becomes available wherever users build and work, grounded in your business logic and governed everywhere.
Connect tools like Cursor and Claude Code directly to Salesforce with native MCPs, APIs, Skills, and CLI commands.
Deploy experiences across every surface. Build once and deliver to Slack, web, mobile, and third-party apps and agents without rebuilding for every channel.
Extend Salesforce with enterprise-grade trust built in. Every experience inherits identity, permissions, compliance, observability and secure real-time data access.
Open. Composable. Trusted.
Inside the developer workflow behind Indeed's production agents, powered by headless capabilities.
Here’s how Salesforce uses headless IT operations to eliminate browser friction and automate engineering workflows.
Decoupling logic from the UI to deploy secure enterprise agents in days, not months.
How a 50-person robotics manufacturer replaced its ERP, saved $200k in fees, and unified every team on one platform.
Health Cloud scales billing faster at a predictable cost with unified patient data.
With headless capabilities, our teams, our teams get the work done directly inside the tools they use everyday, while Salesforce powers the entire business seamlessly in the background.
Andrew RussoVP of Business Systems, BACA Systems
Headless Toolkit is an architectural transformation that exposes every capability Salesforce has built over 25 years directly as an API, a Model Context Protocol (MCP) tool, or a CLI command. It essentially makes the browser interface optional, allowing AI agents to natively plug into Salesforce's data, workflows, and business logic without a human ever needing to open a browser tab. The release includes over 60 new MCP tools, 30+ preconfigured coding skills, and a new Headless Experience Layer to support the entire agentic development lifecycle
The future of enterprise software is agentic, but traditional software was built for humans clicking through interfaces. AI agents cannot navigate user interfaces to get work done. Headless Toolkit bridges this gap by turning the entire platform into an open execution layer, allowing probabilistic AI agents to execute tasks directly at machine speed. Because agents need more than just intelligence. They need context, workflows, and trust. Headless Toolkit ensures that AI models can act on decades of accumulated business logic and security guardrails rather than just manipulating raw data.
If your business is transitioning toward an "Agentic Enterprise" and wants to deploy autonomous AI agents, leveraging Headless Toolkit is highly critical. However, this does not mean your human employees have to stop using Salesforce's traditional interfaces. Lightning components and browser experiences are not going away. Instead, Headless Toolkit provides total architectural flexibility, allowing you to meet your customers and employees wherever they work, whether that is in a traditional browser, through voice, or via an AI agent in Slack.
Total Flexibility to Build Any Way You Want: Developers can use their preferred coding agents (like Claude Code, Cursor, or Codex) and industry-standard UI frameworks (like React) to build directly on top of Salesforce without steep learning curves.
Deploy to Any Surface: The new Headless Experience Layer allows you to build an agent's UI and business logic once, and render it natively across Slack, Microsoft Teams, WhatsApp, and custom web apps without rewriting code for every device.
Enterprise-Grade Trust and Governance: Every agent operates under its own distinct, auditable identity, and inherits all of your existing Salesforce permissions, field-level security, sharing rules, and SLA workflows, ensuring they behave securely and transparently.
Faster Time to Value: By avoiding the need to rebuild complex business logic, compliance controls, or unified data profiles from scratch, companies can deploy production-ready AI agents in a fraction of the time.