Better tools, durable memory, and governed multi-LLM workflows for the agents you already use.

01 · Augment · no key

Better tools for your agent

Helixor Code runs as an MCP server. When your agent needs to find or understand code, it queries the structural Code Index — not grep — and reads and writes stateful TODOs. No key, because the agent supplies the model.

02 · Orchestrate · BYOK

Governed multi-LLM SDLC

For nontrivial or regulated work, Helixor Code drives a workflow from the catalog and routes each step to the right model on your key — Claude triages, a human approves, Codex implements, evidence at every gate.

03 · The shared platform

Persistent & permissioned

Indexes and TODO stacks live server-side, so they persist across sessions and machines and can be shared with permissions. Built on the Helixor engine — works with any LLM, or without one.

The hero pipeline

One governed pipeline, the right model for each step.

sdlc.safe_change_delivery.v1 — checkpointed · retryable 3× · resumable · artifact per step

Workflows are authored and type-checked at design time — no dynamic step generation. We state this plainly, because this audience rewards the candor.

The menu-bar companion

One glance, up and to the right.

Run helixor up and a lightweight companion appears in the macOS menu bar — no window to manage, just a status icon that keeps your local setup honest. Its job is to make the invisible visible: whether the companion is running, healthy, or busy, which MCP transport it's using, and whether it's wired to a trial/cloud or enterprise runtime — before you dispatch an agent.

Because the companion scans your indexes locally and serves structural context instead of whole-repo dumps, the tray turns token savings from a claim into a running meter. It stays deliberately thin — it talks to the local companion for machine status and points to the hosted platform for everything else.