> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mutagent.io/llms.txt
> Use this file to discover all available pages before exploring further.

# FAQ

> Common questions about Helix — scoring, data, the two forms, modes, and removal.

<AccordionGroup>
  <Accordion title="What does a “score” actually look like?">
    Every criterion — derived from your own traces, not invented by us — gets a pass/fail with a
    confidence. Those roll into one verdict on the whole run: a critical failure sinks it. Below a
    threshold of human labels, a judge is marked `unvalidated` rather than trusted, so you always
    know which verdicts to believe.
  </Accordion>

  <Accordion title="Is my data uploaded anywhere?">
    Not to us. Traces stay on your machine; Helix reads them locally. With the binary, model calls go
    from your machine to the provider you signed in with, on your own credentials — we hold no
    provider key and add no inference bill; the plugin reasons on your coding agent's model. Nothing
    is sent to MutagenT unless you ask: `/feedback` shows you what it will send and waits for your
    confirmation, and `helix update` contacts the release server only to fetch a build. The hosted
    Platform is opt-in, through the separate `mutagent` CLI.
  </Accordion>

  <Accordion title="What's the difference between the binary and the plugin?">
    The [**standalone binary**](/helix/install/standalone) is Helix as its own coding agent — one
    install, the whole loop inside, plus the agent and Prime modes, the sub-agent views and the
    trace viewer. The [**plugin**](/helix/install/plugin) installs the Helix skills into the Claude
    Code or Codex you already use. Both run the same loop.
  </Accordion>

  <Accordion title="Which mode should I start with?">
    `helix`. It is the full product — the orchestrator, the lifecycle stages, the sub-agent crew.
    `helix agent` and `helix --prime` are subtractions from it for when you want less; see
    [Three ways to run Helix](/helix/modes).
  </Accordion>

  <Accordion title="Do I need a provider key?">
    For the binary, yes — it talks to the model provider directly. Run `/login` inside the session to
    set up a provider or subscription, or export a key (`ANTHROPIC_API_KEY`, `OPENAI_API_KEY`,
    `GEMINI_API_KEY`, …) before launching. A launch with no provider warns "No models available.
    Use /login to log into a provider via OAuth or API key." — that is the missing key, not a broken
    install. The plugin needs none: it reasons on your coding
    agent's model.
  </Accordion>

  <Accordion title="Do I need traces to start?">
    No. If you're building something new, start at [Spec](/helix/lifecycle/spec) and
    [Build](/helix/lifecycle/build). Evaluate and Diagnose start working once you ship and your
    agent begins logging runs.
  </Accordion>

  <Accordion title="How do I remove it?">
    **Binary:** `curl -fsSL https://install.mutagent.io/helix | bash -s -- uninstall` (add `--purge`
    to drop the cached runtime too). **Plugin:** delete the skill directories under `.claude/` and
    remove the Helix section from your `CLAUDE.md`.
  </Accordion>
</AccordionGroup>