Salesforce unveils Headless 360, turning its entire platform into infrastructure for AI agents

Salesforce Bets Big on AI Agents with Platform Transformation

Salesforce has introduced Headless 360 at its annual TDX developer conference in San Francisco, marking what the company describes as the most ambitious architectural transformation in its 27-year history . The initiative exposes every capability in the Salesforce platform as an API, Model Context Protocol tool, or CLI command, enabling AI agents to operate the entire system without ever opening a browser .

The announcement ships more than 100 new tools and skills immediately available to developers, fundamentally changing how the enterprise software giant approaches application development . Joe Inzerillo, Salesforce president of enterprise and AI technology, explained the vision: "The developer, the builder is talking to these tools, the tools are driving the Salesforce UI creation, configuration etc. We're trying to create this ecosystem where in the future most of the code is going to be written by the agents" 1

1

.

Headless 360 Makes Platform Accessible via APIs and MCP Tools

The Salesforce platform, which includes CRM, customer service, marketing, and ecommerce capabilities, historically required users to interact through web browser interfaces 1

1

. Headless 360 changes that paradigm entirely by making the platform legible to AI agents through standardized protocols 4

4

. Every Salesforce system is now represented through endpoints that agents can use autonomously .

This agentic platform delivers more than 60 new MCP tools and 30-plus preconfigured coding skills that give external coding agents like Claude Code, Cursor, Codex, and Windsurf complete, live access to a customer's entire Salesforce org, including data, workflows, and business logic . Developers no longer need to work inside Salesforce's own IDE, instead directing AI agents from any terminal to build, deploy, and manage applications .

Agentforce Vibes 2.0 Brings Conversational Interface to Development

Agentforce Code, also known as Agentforce Vibes, is a browser-based IDE based on Visual Studio Code now available as part of the free Salesforce Developer Edition 1

1

. The updated Agentforce Vibes 2.0 includes an "open agent harness" supporting both the Anthropic agent SDK and the OpenAI agents SDK, with multi-model support including Claude Sonnet and GPT-5 .

The default LLM in Agentforce Code is Claude Sonnet 4.5, operating in two modes: plan mode and act mode 1

1

. Salesforce has set usage limits for its developer edition at 110 requests per month and 1.5 million tokens, which will refresh monthly until May 31 1

1

. The platform also introduces native React support, allowing developers to build fully functional applications using React rather than Salesforce's own Lightning framework, while inheriting all platform security primitives .

Operating Model Transformation Beyond Product Updates

Industry analysts view Headless 360 as evidence that Salesforce is executing a fundamental operating model transformation rather than simply adding features 4

4

. By exposing capabilities via Headless 360, Salesforce is making them legible to AI agents, effectively reprogramming its operating model rather than its products . The company is re-optimizing its platform for agents, not people, because agents don't need applications or prescribed flows—they consume context, identify tasks, and select capabilities as tools to act .

Madhav Thattai, EVP and GM of Salesforce AI, emphasized the initiative is designed to enable non-programmers to build on the platform: "We want to expand the tent to bring in people that historically have not been part of the Salesforce ecosystem, that now are starting to build capabilities on Salesforce" 1

1

. One enterprise customer saw Agentforce adoption jump from 22% to 78% in six weeks after changing the experience layer 5

5

.

Testing Center and Governance Tools Address Agent Reliability

Recognizing that agents are "probabilistic, not deterministic" and can "reason their way to unexpected outcomes," Salesforce introduced comprehensive governance tools 1

1

. The new Testing Center reveals logic gaps, policy violations, and inconsistent outputs before deployment 5

5

. Custom Scoring Evals let enterprises define what "good" looks like for their specific use case, measuring performance against standards rather than simple pass-fail criteria 5

5

.

Agent Script, Salesforce's domain-specific language for defining agent behavior deterministically, is now generally available and open source . Observability and Session Tracing capabilities help teams identify root causes when agents drift off-task, reducing debugging time from weeks to hours 5

5

. A new A/B Testing API allows multiple agent versions to run against real traffic simultaneously for data-driven optimization 5

5

.

Strategic Response to Enterprise Software Disruption

The timing of this platform transformation comes as Salesforce navigates turbulent conditions in enterprise software markets. The iShares Expanded Tech-Software Sector ETF has declined roughly 28% from its September peak, driven by fears that AI could render traditional SaaS business models obsolete . Salesforce was among companies affected by the SaaSpocalypse scare in February 2

2

.

Inzerillo was explicit about agents crossing organizational boundaries: "I think one of the things about agents is they hate boundaries. Agents do not like to stay in the silo. They want to talk to each other. They want to talk to endpoints" 4

4

. This suggests the challenge is no longer just building agents that navigate internal systems, but making those systems accessible to agents the enterprise doesn't necessarily control . Salesforce also announced a major expansion to Agent Fabric, its trusted agent control plane for companies building multivendor agentic networks, bringing automated discovery, agent authoring, and centralized LLM governance 5

5

.