Automation guide
How to pick an agent's next tool with a decision model
Tool selection is the purest version of the problem: a closed set supplied per call, decided hundreds of times per session, where a generated reasoning trace per step is not affordable.
The decision, typed
Out of the tools this agent has, which one should it call next? As a typed decision this is a choice: one option from a closed set you supply per call, over the tools registered for this agent.
How to actually set this up
Write down the option set before you pick a model
If you cannot enumerate the valid answers, this is not a typed decision yet and no model in the directory will help. The list is an argument to the call, so it can change later, but it has to exist now.
Freeze a test set from decisions you already made
Historical cases plus what was actually decided. Hold it out and never tune against it. Where your own records disagree with each other, keep the disagreement rather than cleaning it away.
Measure accuracy at coverage, not accuracy
A single accuracy figure cannot tell you what to automate. Sort by the model’s confidence, pick thresholds, and read off what share of the queue clears your error budget at each one.
Check the calibration before you trust the threshold
If the model says 0.9 on a hundred cases, roughly ninety should be right. When that does not hold, every threshold you set on top of it is wrong in a way accuracy will not reveal.
Staff the escalation path before you ship the automation
When the agent genuinely needs to reason about why before it acts, and a person will read that reasoning. A decision model returns a label, not an argument.
Models worth shortlisting
Commercially licensed, open weights, small enough to run where the decision happens. None has an independent benchmark, so this is a shortlist rather than a ranking.
laya
The most-liked open-weight decision model of the post-Jev wave, a 421M ModernBERT-large encoder. The base release of a three-model family.
JEV-27B-VL
Qwen3.8-27B fine-tuned to return typed decisions, 28B parameters.
JEV-35B
Qwen3.5-35B-A3B fine-tuned to return typed decisions, 36B parameters.
JEV-27B
Qwen3.8-27B fine-tuned to return typed decisions, 27B parameters.
Narrow it further: size: under 500m · format: gguf
People already doing this
Open-source projects in the agents cluster. Working code to read before you build your own.
yao 8.3k ★
✨ All your agents and workspaces in one place, on every device you own. Track tasks on a board, accessible from desktop, mobile, browser, or API. Self-hosted.
jev-chat-jarvis 8.2k ★
The chat decision assistant: before you reply, Jev reads the chat, judges intent and risk, and drafts replies you fill in with one tap. You press send. Android · Windows · macOS · iOS | 聊天辅助决策工具：先看懂对方再回复，发不发由你。
py-gpt 2.0k ★
Desktop AI Assistant powered by GPT, Gemini, Claude, Grok, Ollama, DeepSeek, Perplexity, and more - chat, agents, tools, MCP, plugins, RAG, vision, voice, image and video generation, speech synthesis and speech recognition, web search, memory, presets, canvas, assistants, themes, computer use. Windows, Linux, and Mac.
agentconnect 1.5k ★
The open-source, multi-agent alternative to Claude Tag. @ any agent, wherever work happens, they work alongside your team, learning as they go.
jev-chat-windows 869 ★
JevChat-Windows：聊天窗口旁挂的回复辅助。窗口截图 + 本地离线 OCR 读对方消息 → Jev 判断意图 → 3 条候选一键填入，发送永远手动
tiptour-macos 690 ★
Open-Source fast local computer use
Other decisions people automate
Which queue does this ticket belong in, and can it be routed without a person looking?
Does this invoice need a human to approve it, or does it match the purchase order closely enough to pass?
How severe is this alert, and does it need to wake someone up?
Did the agent actually complete the step it claims, and was the tool call the right one?
Where do I draw the line between automate and escalate?
Which of these retrieved chunks are actually worth putting in the prompt?
Before you build it
What people ask at this point
The questions that come up once the decision is written down and the option set exists.
Do I need a decision model for this, or will a general LLM do?
A general model behind constrained decoding produces a typed answer too. What it does not reliably give you is a calibrated probability across the option set, and without that you cannot set a threshold. If you are going to send everything to a human anyway, you do not need the calibration and a general model is fine. The longer answer is here.
How much historical data do I need before this is worth trying?
Enough cases with recorded outcomes that you can hold out a test set and still have something left. A few thousand is comfortable. What matters more than volume is that the recorded outcome is what actually happened, not what a workflow defaulted to.
What if my own team disagrees about the right answer?
Score the cases your team agrees on separately from the ones they do not. A single accuracy figure across both hides which kind of error you are actually making, and the contested subset is usually where the model looks worst.
Which model should I start with?
Licence first, then size. Apache-2.0 or MIT, small enough to run where the decision happens, and shipping in a format you can actually deploy. The commercially licensed models are here. None of them has an independent benchmark, so treat the first pick as a candidate rather than an answer.
Find out what your data actually supports
We turn your historical decisions into a frozen test set, then tell you which model, which threshold, and how much of it you can safely automate.