Atomic
Verifiable coding-agent runtime built on checkable execution graphs, listed with an optional Jev provider for bounded choices like routing.
Overview
Atomic is a coding agent runtime that turns an engineering process into explicit, checkable execution graphs. It favours quality over speed: complex runs can take hours or longer because Atomic researches, implements and verifies the work, and you can describe workflows in plain English that it turns into runnable TypeScript graphs. The README reports user claims such as reduced manual verification and a high merge rate.
According to the list summaries, Atomic ships a first-class Jev structured-output provider, so bounded decisions such as routing come back typed through the same decision resolver as its other providers. That integration is not described in the README excerpt available here.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Bounded structured choice in a workflow graph, such as which route to take(inferred) | — |
How it works
- Describe a workflow; Atomic generates the execution graph
- Decision nodes resolve through a provider (optionally Jev)
- Implementation and verification steps run
- Evidence is gathered for review
Strengths
- Typed decisions through a common resolver
- Verification-focused runtime
Caveats
- Jev provider not documented in the README excerpt
- User-reported metrics are not independent
At a glance
- Type
- Service
- Status
- production
- Language
- TypeScript
- Runtime
- coding agent runtime
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Coding agent runtime: ships a first-class Jev structured-output provider so an agent's decisions come back typed, through the same decision resolver as its other providers.
yibie/awesome-jevAn optional Jev decision backend in the Atomic coding Agent for bounded structured choices such as routing.
logicrw/awesome-jev-projects
Frequently asked questions
How is Jev used?
As an optional structured-output provider for bounded choices like routing, per the listing.
What is Atomic for?
Running long, verifiable coding workflows as explicit graphs.
Are the merge-rate numbers verified?
They are user reports quoted in the README.
Related projects
AutoGPT agent platform, listed as adding Jev decision blocks for typed routing, filtering, scoring and confidence-gated next actions.
ProjectproductionPythonAlibaba's AgentScope 2.0 agent framework, which added a JevClassifierModel for binary, choice and score decisions via the TypeSafe SDK.
ProjectproductionPythonLangChain's Python framework, which includes an optional TypeSafeClassifier runnable plus model-routing and risky-tool middleware using Jev.
PluginproductionPythonLangChain.js, which includes an optional TypeSafeClassifier runnable and middleware for bounded routing and tool-call checks with Jev.
PluginproductionTypeScriptOpen-source workspace for building, deploying and monitoring AI agents and workflows, listed as integrating TypeSafe System One.
ProjectproductionTypeScriptGlue that turns one Linux machine into a team of coding agents, with Jev making typed calls like who builds a task and whether it may merge.
ProjectbetaTypeScript