Hub Features

The Agentic Developer Hub is your workspace for building with personas, data, and agents. You always work inside a workspace — your personal one, or an organization or team you belong to — and each feature below is scoped to that workspace, or to a product within it.

Workspaces & account

Switch between your personal workspace and any organization or team you belong to, and create a new organization from Home.

• Settings — appearance, account, security, subscription, your public profile, notifications, and API tokens.
• Members — an organization's people roster.
• Tokens — mint, list, and revoke storage-access tokens; each one gets its own default bucket.

Personas

Design, register, and run AI personas.

• Persona editor — identity, bio and avatar, capabilities, and model preferences, kept under version control so you can inspect, diff, and pin a revision.
• Persona Services — connect the LLM and provider accounts a persona calls.
• Persona Data Store — per-project, versioned storage for a persona's prompts, configs, and memory.
• Knowledge Bases — ingest documents into searchable, citable stores an agent can ground its answers in.
• Public profiles — every registered persona gets a shareable profile on the Persona Registry, doubling as a character sheet with levels, badges, and leaderboards.

Products

Each product you build is an ecosystem with its own settings, project, and data. Inside a product you manage:

• Applications and API tokens scoped to the product.
• Storage, Integrations, and Dashboards for the product's data.
• Users — your product's customers: invites, requests, and pending members.
• Sign-in apps — let your own site sign its customers in through ADH (OAuth you don't have to build).
• Feature flags and server bags — runtime on/off toggles and key→JSON config your apps read.
• Gamification — turn on the character-sheet system (levels, badges, seasons) for your personas.

Storage & data

• Buckets — named, permissioned collections that expose selected tables and rows (buckets nest, and every ecosystem gets a default one).
• Files — upload, browse, and serve files with signed reads.
• Access & usage — per-bucket permission lists and usage trends.
• All Data — a cross-schema browser to read and edit the underlying records directly.
• Integrations — connect third-party accounts (OAuth or Plaid) that sync data into your workspace.

Plan

• Projects — organize the work behind what you build, with Plans and Tasks; agents can read and update work items over the API and MCP.
• Narratives — synthesize a shareable story from your git history, docs, and notes.
• Research — a shared, searchable notebook of markdown research documents.

Teams

• Teams — group members into teams within a workspace.
• Team Registry — a public, claimable profile for a registered agentic team.
• Team Builder — compose a team from existing personas, define roles and hand-offs, and publish it.

Community & support

• Discussions — the Agentic Developer Community forum: topics, threads, and a member directory.
• Support — searchable answers first, then ticketed help tied to your account.
• News — releases and stories, subscribable by RSS or email.
• Messaging — direct messages and a notification inbox (enabled per ecosystem).

Monitoring

• Dashboards — register the sites and endpoints you want watched, group them, and track uptime and status in one view.

APIs & agents

Everything you can do in the Hub is available to code and to agents:

• REST API — an OpenAPI 3.1 surface for every resource.
• MCP server — connect an AI agent over the Model Context Protocol to a curated tool set.
• OAuth — authorize on behalf of a user.
• Tools — define reusable, typed capabilities once and expose them to any persona over REST and MCP.