Features
Multi-model support (DeepSeek V4, GLM 5.2, OpenAI Codex, Anthropic, etc.)
Sub-agent parallel execution with per-sub-agent provider routing
MCP consumption and exposure via stdio and HTTP/SSE
Constitution-based agent identity and authority hierarchy
Plan, Agent, and YOLO modes with orthogonal approval policies
Sandboxed execution: macOS seatbelt, Linux landlock
Lifecycle hooks for pre/post tool execution and mode changes
Public HTTP Runtime API for integrations and automation
Context input budget reuse across routes
Model Lab for testing and comparing models
Fleet/WhaleFlow for durable task execution and fleet management
Community activity feed on homepage
One-line npm install (no Rust toolchain required)
Laguna XS.2 model (33B params, 3B active, on-device)
Laguna M.1 model (225B params, 23B active, via API)
256K context length for multi-step reasoning
Multi-agent orchestration with planning and tool use
Sandboxed agent execution environments
Developer surfaces: agents, TUI, IDE extensions, binaries
Custom foundation models deployed on-prem or in VPC
Data connectors to repos, databases, data warehouses
Role-based access control for humans and agents
Executive-grade governance and auditability
On-prem, VPC, and workstation deployment (defense only)
Forward Deployed Research Engineers embed with teams
Real-time observability with end-to-end traces
Air-gapped network support
Open-weight model weights available
