fstech-digital

fstech-digital / operational-ontology-framework

Public

Framework de Ontologia Operacional — Pin/Spec/Handoff protocol for stateful AI agents. CC BY 4.0.

agent-memory agent-state ai-agents cc-by-4 filesystem-memory
16
4
100% credibility
Found Apr 13, 2026 at 20 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A file-based framework for operating production AI agents with auditable decisions and state tracked in simple markdown documents versioned by git.

How It Works

1
📖 Discover reliable AI helpers

You hear about a clever way to run AI assistants for real work, keeping everything trackable without fancy storage.

2
🗂️ Set up your project

Create a simple folder with ready-made guides for rules and tasks, like starting a new notebook.

3
✏️ Add your rules and to-dos

Jot down the unchanging guidelines for your project and list out what needs doing.

4
🧠 Connect a smart AI

Link up an AI thinking service so your helper can understand and act on instructions.

5
▶️ Launch the helper

Tell it to tackle your first task, and it gets to work right away.

6
✨ Watch it learn and update

See your AI complete jobs, note what it learns, and save a handover for next time automatically.

7
🔄 Run sessions anytime

Come back later, and it picks up exactly where it left off with full history.

🎉 Trustworthy AI companion

Celebrate having a reliable AI that handles ongoing work perfectly with every step recorded safely.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 20 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is operational-ontology-framework?

This Python framework runs stateful AI agents in production using a pin/spec/handoff protocol, storing auditable state in git-versioned markdown files instead of model memory or vector databases. Developers get a CLI tool to boot agents with domain rules (pin), task checklists (spec), and session handoffs, executing cycles that update facts and skills autonomously. It supports Anthropic, OpenAI, and Ollama models out of the box via simple env vars like ADAPTER=ollama.

Why is it gaining traction?

Unlike LangChain or CrewAI, which tie state to vector stores and custom runtimes, this uses plain filesystem + git for zero-infra portability—swap models without losing history, with every decision bisectable via git log. The agent self-writes memory during execution, enabling sober state sovereignty and proven migrations across providers. At 16 GitHub stars, it hooks devs tired of fuzzy embeddings, ranking high for audit-heavy operational ontology needs.

Who should use this?

AI engineers building persistent agents for customer support, business ops, or consulting workflows where decisions need git-tracked justification. Teams migrating LLM providers or running local models on desktop/laptop setups without DB overhead. Not for stateless chatbots or quick prototypes.

Verdict

Grab it under CC BY 4.0 if you need auditable stateful agents—solid docs, CLI, and tests make it production-viable despite 1.0% credibility score and low 16 stars signaling early maturity. Pair with orchestration tools for broader use; skip if vector recall fits better.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.