docyrus-platform

v2026.09.24

Architecture and building-blocks reference for the Docyrus platform — what the platform is made of (apps, data sources, fields, enumerations, custom queries, automations) and how the pieces fit together. Use when you need conceptual context about how Docyrus works, what its building blocks and capabilities are, or to decide which dedicated skill to use for a task. This is the conceptual map and router; hands-on how-to lives in the dedicated design / CLI / API skills it points to (data sources, automations, agents, AI tools, templates, webforms, ACL, querying, the CLI, and the REST API).

GitHub
安装命令
npx skhub add docyrus/docyrus-platform
Markdown
SKILL.md

Docyrus Platform

Docyrus is an AI-native Backend Platform as a Service (BPaaS) for building B2B web/mobile apps, client portals, internal tools, AI agents, chatbots, and integrations — without building backend infrastructure from scratch. This skill is the architecture map: the building blocks, how they fit, and which dedicated skill owns each how-to.

Core Building Blocks

The platform is composed of building blocks that compose into apps:

  • Apps — Top-level containers that group data sources, automations, agents, and custom queries into a deployable unit.
  • Data Sources — Structured collections of records with typed schemas. Four types: simple, advanced, external (connected DBs/APIs), and system (pre-built). Every data source — internal or external — is exposed through one unified CRUD endpoint.
  • Fields — 45+ field types defining each data source's schema: text, numbers, dates, selections, relations, files, formulas, nested data, and more.
  • Enumerations — Reusable option sets for selection fields (color, icon, ordering).
  • Custom Queries — SQL-based analytics templates with variable interpolation, runtime filters, and multi-database targeting.
  • Automations — Event-driven workflows: triggers (record changes, schedule, webhook, email, webform, button) → action-node chains (email, notifications, record ops, HTTP, AI, scripts).

Detailed building-block specifications: references/core-building-blocks.md.

Which skill for which task

This skill holds the concepts. For doing the work, use the dedicated skill:

To…Use skill
Design a data source (schema, fields, relations, enums, access)docyrus-data-source-design
Build an automation (triggers + action-node graph)docyrus-automation-design
Build a custom AI agent + its sub-resourcesdocyrus-agent-design
Author AI tools the system base agent ("Docy") callsdocyrus-app-ai-tools
Design an email templatedocyrus-email-template-design
Design an HTML/PDF print/report templatedocyrus-print-pdf-template-design
Build a public webform that collects submissions into recordsdocyrus-webform-design
Design a data source's record create/edit/view form (grid, sections, conditional rules)docyrus-dynamic-form-design
Configure tenant ACL (roles, hierarchy units, role rules)docyrus-acl-design
Discover & call third-party integrations/connectors (docyrus connect)docyrus-integrations-and-connectors
Query/CRUD data, write filters/formulas, build query payloadsdocyrus-api-dev
Write DSQL (logical SQL) queries, or natural-language dsql askdocyrus-dsql-query-design
Drive everything from the terminal (auth, ds, studio, automation, agent, …)docyrus-cli-app
Build a React app on Docyrus (collections, grids, forms)docyrus-app-dev-react, docyrus-data-grid-page-design, docyrus-record-detail-form-design

Capability areas (concepts)

Querying & Data Operations

Unified query engine: column selection, 50+ filter operators, aggregations, formulas, pivots, child queries, full-text search; full CRUD with bulk ops, comments, and attachments; plus a read-only logical SQL endpoint (DSQL) over appSlug.dataSourceSlug tables. Concept overview: references/querying-and-data-operations.md. → Query payloads & formulas: docyrus-api-dev. DSQL (logical SQL, dsql ask): docyrus-dsql-query-design. Terminal querying: docyrus-cli-app.

AI Capabilities

AI agent builder with tool binding, knowledge bases, MCP servers, 18+ model providers, agent chaining, task scheduling, persistent memory, chat integrations, and evaluation metrics. Concept overview: references/ai-capabilities.md. → Build agents: docyrus-agent-design. Author the tools Docy calls: docyrus-app-ai-tools.

Automation & Workflows

Event-driven automation engine: trigger types and action-node types, conditional flows, action chains, archiving. Concept overview: references/automation-and-workflows.md. → Build automations: docyrus-automation-design.

Authentication & Multi-Tenancy

OAuth2 flows (PKCE, Client Credentials, Device Code), scope-based permissions, tenant isolation, role-based and record-level ACL, React auth helpers, and the client portal. Concept overview: references/auth-and-multi-tenancy.md. → Configure ACL: docyrus-acl-design. OAuth in app code: docyrus-api-dev.

Integrations & Events

Connector framework for HTTP and SQL providers, webhook/event management, connector discovery, collaborative document editing, in-app messaging, notifications, and email. Concept overview: references/integrations-and-events.md. → Discover & use connectors (docyrus connect): docyrus-integrations-and-connectors. Webhook/messaging CLI: docyrus-cli-app.

Platform Services

App templates, data import/export, reporting & analytics, deployment & versioning, localization & navigation, audit logging, and billing. Concept overview: references/platform-services.md. → Public data-collection webforms are their own skill: docyrus-webform-design.

Developer Tools

Auto-generated OpenAPI specs, MCP server, a full-featured CLI, REST API client libraries, a React auth provider, and auto-generated collection hooks. Concept overview: references/developer-tools.md. → Full CLI command index: docyrus-cli-app (docyrus <cmd> --help for flags). REST client + query payloads + formulas: docyrus-api-dev.

发现
标签

此技能尚未发布标签。

版本
最新版本元数据

版本

v2026.09.24

发布时间

Sep 24, 2026

分类

未分类

许可证

未指定

源路径

skills/docyrus-platform

默认分支

main

最新提交

5eb990f

Tree SHA

d012754