Hi Product Hunt 👋 I’m Jeremy, and today I’m launching AgentCannabis, an open-source library of portable AI skills built for the Canadian cannabis industry. Cannabis is an interesting domain for AI because the challenge is not simply giving the model more information. Real work involves cultivation records, quality systems, post-harvest controls, inventory, processing, regulatory evidence, source currency, professional responsibility, and decisions that an AI system should never pretend it has the authority to make. AgentCannabis was built around that distinction. Instead of one large cannabis prompt, the project breaks the domain into structured, reviewable skills that help AI systems examine records, identify missing evidence, preserve uncertainty, and prepare clear handoffs for the responsible human or qualified professional. The current release includes: 🌱 233 atomic skills across 18 taxonomy families, spanning cultivation, plant health, drying and curing, processing, quality, inventory, compliance, and operational governance 👥 18 professional skillsets for roles including cultivation technicians, master grower support, quality systems, QAP support, processing, post-harvest, inventory, preventive controls, and compliance 🇨🇦 Canadian-focused regulatory architecture designed to preserve licence, role, activity, site, product, source, and approval boundaries 📋 Evidence-first workflows for reviewing records, identifying gaps, checking source currentness, documenting uncertainty, and preparing responsible-human handoffs 🧪 Deterministic validation and evaluation tooling covering repository structure, source links, review gates, scenarios, adversarial checks, and public distribution One design principle matters throughout the project: Review assistance is not authorization. AgentCannabis does not approve production, release cannabis lots, submit regulatory filings, authorize pesticide use, operate extraction equipment, sign records, or replace qualified legal, regulatory, quality, engineering, or operational review. The goal is to make AI more useful inside regulated cannabis workflows without making it less accountable. The project is MIT licensed, open source, and currently released as v1.0.5. I’d particularly like feedback from licensed producers, growers, quality teams, compliance professionals, processors, and AI developers: Which cannabis workflow still depends most heavily on manually reviewing records, evidence, and handoffs, and where could AI realistically remove work without taking authority away from the responsible professional? GitHub: https://github.com/jeremylongwor...

AgentCannabis AI Agent Skillsets

Practical AI Skills for Cannabis Cultivation and Refining

Practical AI Skills for Cannabis Cultivation and Refining