I built Khaos Brain, an open-source local-first experience system for AI agents

Ingo — Wed, 29 Apr 2026 18:03:12 +0000

I built Khaos Brain because most AI memory features feel too shallow for real agent work.

Saving "remember this next time" is useful, but the more valuable unit is accumulated experience: what condition appeared, what action was taken, what result happened, which route failed, and which route later became reliable.

The problem is not that an agent cannot remember a sentence. The problem is that after doing similar work many times, its working experience often does not accumulate into something inspectable and reusable.

Khaos Brain is an open-source, local-first experience organization tool for AI agents. It stores experience as visible file-based cards instead of opaque memory.

What it does

The current release is Codex-first, but the idea is broader:

before a task, the agent can retrieve relevant experience
after a task, it can write back observations and lessons
maintenance workflows can organize those cards over time

The cards can be searched, inspected, reviewed, diffed, merged, and rolled back with Git.

One part I care about is keeping personal memory and shared knowledge separate. Personal preferences stay local. Reusable task models, engineering lessons, and reviewed skills can be shared through a GitHub-backed organization knowledge base.

The easiest way to try it is to hand the GitHub URL to your coding agent and ask it to install and enable the project.

Repo:
https://github.com/liuyingxuvka/Khaos-Brain

Feedback:
https://github.com/liuyingxuvka/Khaos-Brain/discussions/2

I would especially like feedback on:

whether "local-first experience system" works well as a framing
whether visible experience cards feel more useful than opaque AI memory
what kind of agent work would benefit most from accumulated experience

I built Flow-Guard, an open-source finite-state workflow simulator for AI coding agents

Ingo — Wed, 29 Apr 2026 18:02:21 +0000

I built Flow-Guard because I keep seeing the same failure mode with AI coding agents: a local change looks correct, but the larger workflow or architecture is still wrong.

Flow-Guard is an open-source Python tool that acts like a lightweight workflow / architecture simulator before an agent writes or changes code.

It is not an LLM wrapper and it is not prompt-only. The core is a real finite-state mathematical model: you define the workflow state, transitions, inputs, and outputs, then Flow-Guard explores that model and reports concrete counterexample traces.

In other words, it is using a state-machine style mathematical simulation to reason about the workflow before code is generated or changed.

What it tries to catch

Right now I am focusing on three broad classes of problems:

duplicate side effects or repeated actions
state / cache / source-of-truth drift
stuck loops, retry paths, or progress failures

The goal is not to replace tests or claim full formal verification. I am trying to add a design-time guardrail: simulate the logic first, inspect the failures, then let the coding agent implement or modify the real code.

The easiest way to try it is to hand the GitHub URL to your coding agent and ask it to install the project and run the examples.

Repo:
https://github.com/liuyingxuvka/FlowGuard

Feedback:
https://github.com/liuyingxuvka/FlowGuard/discussions/1

I would especially like feedback on:

whether the concept is understandable
whether "workflow / architecture simulator" works well
what real AI-agent workflow bugs would be worth modeling next

DEV Community: Ingo

I built Khaos Brain, an open-source local-first experience system for AI agents

What it does

I built Flow-Guard, an open-source finite-state workflow simulator for AI coding agents

What it tries to catch