Developer Resources
REST API, MCP tools, and building your own apps on the platform.
Portablemind is built on an open, API-first platform. Everything you can do in the app — and a great deal more — is available programmatically, so you can integrate Portablemind with your own systems or build entirely custom applications on top of it.
This documentation site covers using the Portablemind app. Developer-level material lives in the platform API guide, summarized below.
The API guide
The API guide is the home for everything code-level:
- REST API reference — authentication (JWT + tenant headers), response formats, error handling, and the full set of endpoint groups: parties and contacts, business events, security, and AI.
- MCP tools reference — the Model Context Protocol tools that let AI agents (including external agents like Claude) operate on the platform: creating records, querying data, managing conversations, and more.
- Dynamic Functions — serverless JavaScript that runs in a sandboxed environment inside your workspace, with OAuth integration and scheduling.
- Building your own app — scaffolding a single-page application against the platform, including shared authentication components and working sample code.
- Sample applications — a gallery of prebuilt demo apps showing platform patterns in action.
Interactive API reference
A browsable Swagger/OpenAPI reference is available at www.dsiloed.com/docs/api/v1, covering request and response schemas for every public endpoint.
Where the two guides meet
Some Portablemind features have both a user-facing side (documented here) and a developer side (documented in the API guide):
| Feature | User docs | Developer topics in the API guide |
|---|---|---|
| Conversations & AI | Chat & the AI Assistant | Webhook ingestion, embedding the chat component |
| Guest users | Guest Users | Capability creation and assignment APIs |
| Workspace sharing | Workspace Sharing | Cross-tenant access request APIs |
| Coding agents | Coding Agents (SiloLink) | SiloLink session APIs and MCP bridge |
| Agents | Building AI Agents | Agent and model configuration APIs |
| Dynamic functions | Dynamic Functions | Function management and execution APIs |
Tip: If you're building AI agents that act on the platform, start with the MCP tools reference — it's usually a faster path than the raw REST API.