Back to projects
Active Started Jun 2026

RossLabs Agent Harness

Model-agnostic, local-first Rust agent harness: turns any local or cloud model into a coding agent with deterministic gates, routing, and sandboxed execution.

Private Repo
Rust Ollama MLX llama.cpp OpenRouter

A portable, CLI-first agent harness that treats the model as a swappable data row — local (Ollama / MLX / llama.cpp) or cloud (OpenAI, Groq, Fireworks, OpenRouter, Together, DeepInfra). The value isn’t the model; it’s the deterministic scaffolding around it, encoded in Rust so the model never computes policy:

  • A verification gate that makes cheap/open models safe — on an OSS ingest sweep it rejected 6 of 7 garbage outputs a raw model call would have returned.
  • Routing to the cheapest model that passes each task’s oracle, and a decision on whether to invoke the harness at all vs a direct/deterministic solve.
  • Per-model / per-domain profiles, a Program-Aided code_exec lane, and sandboxed execution with graceful fallback (Docker → macOS Seatbelt → Linux Bubblewrap → explicit opt-in), never running unsandboxed silently.

Self-hosting: once the core loop is reliable on a local model, the harness builds its own later phases.