Local-first · macOS · open source

Build, specialize, and prove expert agents.

RunEvals is a native studio for giving a small model one bounded job, generating the post-training data and evals to teach it, composing agents into hybrids, and freezing a deterministic gate before anything ships. It runs on your machine. No model calls, no network, no repository writes until you make that boundary explicit.

Press ⌘⇧P for anything. Deterministic gates. Ships as an Agent Plugin.

Why a studio

Post-training a specialist should not need a research team.

Most useful agent work is repeated and bounded: triage this, classify that, propose the narrow action. RunEvals turns that into a contract you can inspect: what the agent may touch, what good looks like, and what it may never do on its own.

Scoped authority

Every tool is read-only, human-approval, or production-write. Production stays locked until a real quality gate passes and a person approves.

Deterministic gates

Fail-closed checks catch missing scope, thin data, no held-out set, or an unlocked write, before a single token is spent.

Portable contract

Everything is one versioned .runevals.json. Open it, diff it, hand it off, or publish it.

The taller layer

Generate the data and the evals, not just the agent.

A package format ships a capability. It never tells you whether the capability is any good. RunEvals is where you make it good: expand a few seed decisions into a diverse, leak-free dataset, freeze the adversarial cases as gates, and prove a small model replaced the frontier one at a known token budget.

Smart data

Grow seed decisions into normal, edge, recovery, and adversarial cases structured around the task's real failure modes, with clean train / valid / test splits.

Evals that hold

Held-out and adversarial cases become frozen gates. Deterministic, reproducible, auditable. An open way to measure capability, not a closed lab.

Low-token by design

Model the work as a deterministic flow; spend model tokens only at the nodes that need judgement. The per-run budget is visible and small.

Marketplace + hybrids

Borrow an agent. Compose two. Make it yours.

Every agent is a portable contract that carries its own gates. Adopt one, or tick several and compose them into a hybrid: a union of tools, examples, and flow, with production writes still locked. You inherit a re-runnable test, not a claim.

Email triage
READ-ONLY · SORT ON APPROVAL
Receipt extractor
READ-ONLY
PR summarizer
READ-ONLY · PROPOSE
Meeting notes
READ-ONLY · PROPOSE
Where it fits

Standards ship agents. RunEvals proves them.

Open plugin standards solve distribution: a manifest, some skills, an MCP server, portable across clients. They deliberately say nothing about quality, specialization, or measurement. RunEvals owns that layer, and exports to the standard so your work stays shippable.

A plugin standard gives you

  • A manifest and a folder layout
  • Skills and an MCP server, packaged once
  • Discovery across compatible clients
  • Nothing about whether it works

RunEvals adds

  • Generated post-training data and frozen evals
  • A small specialist, proven against the frontier
  • Scoped authority and a token budget
  • Export to the plugin standard, so it still ships
Get it

Runs on Apple Silicon. Builds without Xcode.

Clone, verify, and package a signed local build. Verification is framework-free, so you do not need a full Xcode install.

git clone https://github.com/thehimalayanleo/runevals-mac.git
cd runevals-mac
./scripts/test.sh # deterministic self-check
./scripts/package_app.sh # build + sign RunEvals.app
open dist/RunEvals.app

Requires macOS 14 or newer on Apple Silicon. The local build is ad-hoc signed, not notarized for distribution.