How it works
How rightmodeler works
Detect, prove, fix. The whole loop, run on the traces you already have.
TL;DR
rightmodeler replays your real agent traces through cheaper models, judges each output against what you already shipped, and applies the safe swaps in your repo.
in ▸ raw traces
- 01
Detect
Point it at the traces you already emit. rightmodeler autodetects the format across eight sources and folds every run into one per-step schema, with no new SDK and no re-instrumentation.
reads Claude Code · Codex · LangSmith / LangGraph · OpenAI SDK · Langfuse · Braintrust · Phoenix · OTel GenAI → 1 per-step schema
- 02
Prove
It replays each step through cheaper candidates on your real inputs and judges every output against what you already shipped. Each swap gets a save %, a quality score, the evidence behind it, and a risk flag, and it abstains when the evidence is weak.
scores save % · quality · evidence · risk flag → verdict + confidence · abstain on high-risk
- 03
Fix
You approve each safe swap and rightmodeler applies it in your repo via the Skill. A report and an edit, never a live intercept. You decide what to swap, and when.
applies approved swap in your repo, via the Skill · nothing changes without your approval
out ▸ signed report
Not observability. Not a runtime gateway.
Observability only shows you problems; a gateway hijacks live traffic. rightmodeler proves the fix on runs you already shipped, then applies it in your repo.
FAQ
- Which traces are supported?
- Eight formats, autodetected: Claude Code, Codex, LangSmith / LangGraph, the OpenAI SDK, Langfuse, Braintrust, Phoenix, and OpenTelemetry GenAI. rightmodeler folds them all into one per-step schema, so you point it at the traces you already emit, with no new instrumentation.
- Does it touch production?
- No. rightmodeler replays your past traces offline and produces a report plus a repo edit. It never sits in your request path, routes live traffic, or adds latency. It is not a runtime gateway.
- Do you store my data?
- It runs locally on your own traces and your own API keys. Replays call whichever model providers you choose (directly or via OpenRouter) using your keys; there is no rightmodeler server holding your traces.