Skip to main content

​
Product

OrgX is an operating record for AI-delivered work. It connects initiatives, decisions, artifacts, approvals, and execution receipts so supported AI clients and people can work from the same workspace context.
OrgX is designed for solo technical founders, small teams, and product or engineering leaders who need agent work to remain reviewable across tools. The best fit is work that has meaningful handoffs, approvals, or evidence requirements.
An AI client provides the interaction surface. OrgX provides the shared workspace record: work hierarchy, decisions, artifacts, access boundaries, and receipts. Contract principles explain what state and evidence mean across those surfaces.
The published REST operations and MCP surfaces are documented for real integrations. Use the contract and availability pages for the behavior of each surface, keep authenticated tests in a dedicated workspace, and review consequential actions before enabling broader autonomy.

​
Agents and control

Depending on the connected surface and granted permissions, agents can read context, analyze work, draft artifacts, create or update connected-tool records, and propose decisions. The active MCP tool catalog and integration guide are authoritative for what is callable in a given client.
Workspace authorization, integration scopes, autonomy settings, and decision checkpoints control what agents may do. A producer receipt is not proof of acceptance or an external outcome; review records and evidence remain separate states.
Use workspace autonomy settings, tool scopes, budgets, and supported skill or plugin configuration. See Decisions and Skills for the surfaces available to your client.

​
Integrations

Linear is documented as an available integration. GitHub is documented as a beta integration. Stripe is documented for billing and the Content Studio callback boundary. Slack, Notion, HubSpot, Figma, and Mailchimp are not native OrgX integrations in the current guide. See Integrations for scope and permissions.
Yes. Start with the web app, AI Toolkit, OpenClaw, ChatGPT where remote MCP is available, or the hosted MCP endpoint. Add external integrations only when their documented permissions match your workflow.
Use Settings → Integrations → Disconnect in the OrgX application. This removes OrgX’s access; it does not imply deletion of historical records already stored in the workspace.

​
ChatGPT and MCP

Use the ChatGPT integration guide. It documents the current Developer Mode or custom-connector path for accounts where remote MCP is available. A working private connection is distinct from publication in a provider directory.
The Model Context Protocol is the interface used by compatible AI clients to call OrgX tools and read OrgX resources. See the MCP protocol reference for endpoint, authorization, and transport details.
Yes, when the client supports the documented MCP transport. Start with the AI Toolkit guide or the MCP client setup guide.

​
Billing and privacy

See the current pricing page. This documentation does not repeat plan amounts because pricing and entitlements can change independently of the API contract.
Read Security and the application’s Privacy Policy. Do not put credentials, payment details, or private workspace data in public validation requests or support tickets.