DecisionAct v0.1.2 ships reliable multi-tool calls for LLM-free agents built on decision models
AgitatedReveal07 · reddit · 2026-10-26
DecisionAct is an open-source Python framework that runs task-focused agents on decision models like JEV and Cloudflare Clef instead of LLMs: you define Python tools, inputs, intents, and reply templates; the decision model picks actions and inputs.
v0.1.2 improvements:
- Repeated tool calls now resolve fresh inputs instead of reusing earlier arguments
- Earlier tool results can feed later calls (e.g. "add 4 and 5, then multiply the result by 2")
- Clearer execution context separating instructions, messages, and completed calls
- Uncertain input selections get one recheck, then the framework asks the user
- Longer capped tool sequences and better continuation after human approval
Input resolution uses regex candidates, categorical choices, or text candidates selected via parameter descriptions. Install with pip install -U decisionact; docs include calculator, order-support, and flight-desk examples.
More from coding & agent
- Someone sent the first customer fax via ChatGPT over Onetime Fax's new MCP server — tdhopper · 2026-10-26
- China Telecom's XingChen lab open-sources Xing4.0-29B-Intent, a 29B MoE with 4B activated params — FellMentKE · 2026-10-26
- Codebases sized to fit one context window dramatically improve agent outcomes — julianharris · 2026-10-26
- 42-Minute Video Breaks Down Why AI Agents Fail: Mostly Not the Model's Fault — JohnSavill · 2026-10-26
- Giraffe: Open-Source Baseline Testing for OpenAI-Compatible LLM Endpoints — ImmediateBig8479 · 2026-10-26
- Running a bed-side AI agent: Obsidian split-screen plus Hyo plugin as a mobile workstation — evielync · 2026-10-26