Capability index

Features

What Primitive can do, how the capabilities connect, and how much of each capability is actually implemented. Every page starts simple and keeps the technical depth underneath it.

Status is part of the documentation. “Specified” and “experimental” ideas are intentionally visible, but they are not presented as shipped product behavior.

Work

2 capabilities
WorkPartial

Spaces, Projects & Tasks

Organise work from broad context down to executable units.

Read from TL;DR → technical detail
WorkPartial

Conversations & Comms

Keep communication attached to the work and durable outcomes it creates.

Read from TL;DR → technical detail

AI

5 capabilities
AIImplemented

AI

Use governed AI inside the same workspace as the work and context.

Read from TL;DR → technical detail
AIImplemented

Agents

AI workers with durable runs, scoped context, tools and governed writes.

Read from TL;DR → technical detail
AISpecified

Orchestration & Holarchy

Recursively coordinate work without depending on permanently fixed agent seats.

Read from TL;DR → technical detail
AIPartial

Skills & Expert Routing

Route a problem to the smallest useful set of reusable expertise.

Read from TL;DR → technical detail
AIImplemented

Models, Providers & Routing

Choose execution strategies and providers without making the provider the architecture.

Read from TL;DR → technical detail

Knowledge

4 capabilities
KnowledgeImplemented

Memory & Context

Give work durable context without making every memory globally visible.

Read from TL;DR → technical detail
KnowledgePartial

Knowledge & Search

Store, index and retrieve information with permissions and provenance intact.

Read from TL;DR → technical detail
KnowledgePartial

Capture & Intake

Turn incoming material into traceable candidates for action, knowledge or review.

Read from TL;DR → technical detail
KnowledgeImplemented

Opportunity Intelligence

Turn market and opportunity signals into evidence-bounded analysis.

Read from TL;DR → technical detail

Automate

2 capabilities
AutomateSpecified

Schedules & Automations

Start governed work from time, events or changing conditions.

Read from TL;DR → technical detail
AutomateSpecified

Workflows, Hooks & Webhooks

Connect lifecycle events to deterministic or agentic actions.

Read from TL;DR → technical detail

Connect

2 capabilities
ConnectPartial

MCP & Connectors

Expose tools and external systems through governed invocation boundaries.

Read from TL;DR → technical detail
ConnectPartial

Browser

Give governed agents a browser surface and bring capture back into Primitive.

Read from TL;DR → technical detail

Build

2 capabilities
BuildImplemented

Builder / Create

Create React interfaces with revisions, components, tokens and isolated previews.

Read from TL;DR → technical detail
BuildImplemented

Harness & Execution Runtime

A governed execution boundary between intent, models, tools and effects.

Read from TL;DR → technical detail

Govern

2 capabilities
GovernPartial

Control & Governance

Separate the ability to reason from the authority to create real effects.

Read from TL;DR → technical detail
GovernPartial

Evidence & Provenance

Keep claims, decisions and actions traceable back to their sources.

Read from TL;DR → technical detail