Features
In-process virtual bash interpreter (no OS processes spawned)
164 reimplemented commands (grep, sed, awk, jq, curl, tar, find, xargs, etc.)
POSIX shell language coverage with bash extensions (arrays, [[ ]], brace expansion, coprocesses, traps)
Virtual filesystem (InMemoryFs, OverlayFs, MountableFs) with host mount opt-in
Resource limits (commands, loops, output, input, filesystem size, parser timeout, AST depth)
Default-deny networking with allowlist and SSRF protection
Credential injection and Ed25519 request signing (RFC 9421)
Built-in runtimes: Python (Monty), TypeScript (ZapCode), SQLite (Turso), SSH, Git
LLM tool contract (BashTool) with discovery metadata, streaming output, and system prompts
Snapshotting for checkpoint/resume
Scripted tool orchestration via ToolDef + callbacks
Custom builtins in Rust or JavaScript
Live mount hot-swapping on running interpreter
Structured logging with tracing and redaction
Interactive REPL with line editing
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
