sim

Provides a drag-and-drop canvas for building agent workflows with real-time multi-user collaboration using operational transformation or CRDT-based state synchronization. The canvas supports block placement, connection routing, and automatic layout algorithms that prevent node overlap while maintaining visual hierarchy. Changes are persisted to a database and broadcast to all connected clients via WebSocket, with conflict resolution and undo/redo stacks maintained per user session.

Abstracts OpenAI, Anthropic, DeepSeek, Gemini, and other LLM providers through a unified provider system that normalizes model capabilities, streaming responses, and tool/function calling schemas. The system maintains a model registry with metadata about context windows, cost per token, and supported features, then translates tool definitions into provider-specific formats (OpenAI function calling vs Anthropic tool_use vs native MCP). Streaming responses are buffered and re-emitted in a normalized format, with automatic fallback to non-streaming if provider doesn't support it.

Integrates OAuth 2.0 flows for external services (GitHub, Google, Slack, etc.) with automatic token refresh and credential caching. When a workflow needs to access a user's GitHub account, for example, the system initiates an OAuth flow, stores the refresh token securely, and automatically refreshes the access token before expiration. The system supports multiple OAuth providers with provider-specific scopes and permissions, and tracks which users have authorized which services.

Allows workflows to be scheduled for execution at specific times or intervals using cron expressions (e.g., '0 9 * * MON' for 9 AM every Monday). The scheduler maintains a job queue and executes workflows at the specified times, with support for timezone-aware scheduling. Failed executions can be configured to retry with exponential backoff, and execution history is tracked with timestamps and results.

Manages multi-tenant workspaces where teams can collaborate on workflows with role-based access control (RBAC). Roles define permissions for actions like creating workflows, deploying to production, managing credentials, and inviting users. The system supports organization-level settings (branding, SSO configuration, billing) and workspace-level settings (members, roles, integrations). User invitations are sent via email with expiring links, and access can be revoked instantly.

Allows workflows to be exported in multiple formats (JSON, YAML, OpenAPI) and imported from external sources. The export system serializes the workflow definition, block configurations, and metadata into a portable format. The import system parses the format, validates the workflow definition, and creates a new workflow or updates an existing one. Format conversion enables workflows to be shared across different platforms or integrated with external tools.

Enables agents to communicate with each other via a standardized protocol, allowing one agent to invoke another agent as a tool or service. The A2A protocol defines message formats, request/response handling, and error propagation between agents. Agents can be discovered via a registry, and communication can be authenticated and rate-limited. This enables complex multi-agent systems where agents specialize in different tasks and coordinate their work.

Implements a hierarchical block registry system where each block type (Agent, Tool, Connector, Loop, Conditional) has a handler that defines its execution logic, input/output schema, and configuration UI. Tools are registered with parameter schemas that are dynamically enriched with metadata (descriptions, validation rules, examples) and can be protected with permissions to restrict who can execute them. The system supports custom tool creation via MCP (Model Context Protocol) integration, allowing external tools to be registered without modifying core code.

Executes workflows as directed acyclic graphs (DAGs) with support for loops (for-each, while), parallel branches, and nested workflow calls. The engine maintains execution state (variables, loop counters, branch results) across all blocks, with checkpointing at each block boundary to enable run-from-block debugging. Execution can be paused at human-in-the-loop blocks, resumed from that point, or stepped through one block at a time. Background execution is handled via a job queue (likely Bull or similar) that persists execution state and allows long-running workflows to survive process restarts.

Provides a knowledge base system where documents can be uploaded, chunked, embedded using a configurable embedding model, and stored in a vector database. During workflow execution, agents can retrieve relevant documents via semantic search (cosine similarity on embeddings) and inject them into the LLM context. The system supports multiple embedding providers (OpenAI, Anthropic, local models) and vector stores (likely Pinecone, Weaviate, or PostgreSQL pgvector). Retrieved documents are ranked by relevance score and can be filtered by metadata (source, date, tags).

Allows workflows to be triggered via HTTP webhooks with configurable authentication (API key, OAuth, HMAC signature verification). The system includes webhook deduplication logic that prevents duplicate executions from retried webhook deliveries (using idempotency keys or request hashing). Webhook payloads are parsed, validated against a schema, and injected into the workflow as initial variables. The system supports provider-specific webhook subscriptions (e.g., GitHub push events, Stripe charges) with automatic payload transformation.

Provides an in-app AI assistant that understands the current workflow context (blocks, connections, variables) and suggests next steps, tool recommendations, or workflow optimizations. The Copilot uses the workflow definition as context, maintains a message history with checkpoints for stream resumption, and can execute tools (like searching the block registry or documentation) to provide informed suggestions. It supports multiple modes (chat, command palette, inline suggestions) and can be invoked via keyboard shortcuts or UI buttons.

Manages workflow deployment across multiple environments (dev, staging, production) with version control, rollback capability, and environment-specific variable overrides. Each deployment creates a snapshot of the workflow definition and configuration, which can be promoted through environments or rolled back to a previous version. The system tracks deployment history with timestamps and user attribution, and supports blue-green deployments where new versions run in parallel before traffic is switched.

Captures detailed execution logs for each workflow run, including block-by-block execution trace, variable state at each step, tool call arguments and results, and LLM prompts/responses. Logs are streamed in real-time to the UI via WebSocket, allowing users to watch execution as it happens. The terminal view shows formatted output with syntax highlighting for code, JSON, and logs, and supports filtering/searching logs by block name, timestamp, or log level.

Manages API keys and credentials for external services (LLM providers, tools, connectors) with support for both platform-managed keys and user-provided keys (BYOK). Keys are encrypted at rest and in transit, with role-based access control to restrict which users/workflows can access which credentials. The system supports credential sets that group related keys (e.g., all Stripe credentials) and fan-out to multiple workflows. Credentials can be rotated without updating workflows, and usage is tracked for audit purposes.