Story How it works Tests Insights Whitepaper What's next Download Author Contact

Insights

Ideas, tutorials, and technical deep-dives on AI memory.

Reflection

My AI memory has senses now — so I stopped taking notes.

I used to tell EIDARA what to remember. Now it listens, reads what I type, and notices what’s on my screen — and decides for itself. What changed in practice.

Read more →
Reflection

The stabilization weeks: making EIDARA boring to use.

Two weeks after the senses went live, the real work began — replacing every rough edge with a structural fix that won’t break the same way twice. An update on where EIDARA is.

Read more →
Reflection

I gave my AI three senses. And then I gave myself the off switch.

I gave my AI memory three senses — voice, typing, and screen text. The most important feature isn’t any of them. It’s the off switch.

Read more →
Technical

Why my AI doesn’t have eyes.

I built an AI memory that reads my screen without looking at it. Why I rejected computer vision for the accessibility tree.

Read more →
Technical

When two AIs argue about my memory.

Two AIs disagreed about a fact in my memory. I built four ways for that conflict to resolve — and only one of them involves me.

Read more →
Reflection

The day I asked my system to grade itself.

An outside agent audited my memory system and found five debts the dashboard wasn’t showing. The dashboard was lying.

Read more →
Reflection

The day my system stopped crying wolf.

The auditor reported the same issues daily that the curator was already fixing. The principle: every failure becomes a prevention rule, not a patch.

Read more →
Technical

The neuron that became a garbage magnet.

One memory neuron silently accumulated 156 facts that didn’t belong. A cascade misrouting bug, and how a single guard fixed it overnight.

Read more →
Technical

Two weeks of autonomous AI memory: what actually happened.

Real data from two weeks of autonomous AI memory. What worked, what broke, what we changed. No theory — just numbers and decisions.

Read more →
Technical

Any AI can now feed my memory. None of them touch it.

Multiple AIs contribute atomic facts via MCP. One local curator decides what gets integrated. The folder is dead — DARA is a service now.

Read more →
Reflection

The day my AI memory stopped needing me.

I activated an autonomous curator. Five hours later, five false positives. Two weeks later, zero errors. Here's what happened.

Read more →
Reflection

One week in: notes from a self-tending memory.

140 cycles, zero errors. A sleep-time compute agent curating local-first AI memory — what happened, what broke, and what's next.

Read more →
Reflection

The first day of a self-tending brain.

We activated the autonomous side of DARA. Five hours later, five false positives. Here's what we did about it.

Read more →
Reflection

EIDARA v2.0 is here — and the next direction.

v2.0 makes the system self-correcting. Plus: atomic memory and desktop context.

Read more →
Reflection

I stopped organizing my AI's memory into folders.

Why I abandoned hierarchical folders for a compiled, flat approach to AI memory.

Read more →
Reflection

Why I don't use specialized AI agents anymore.

One AI with full context outperforms six agents without it.

Read more →
Reflection

The MECE trap: how Bain taught me to think, but not to remember.

MECE is a thinking tool, not a memory tool. Here's why I had to unlearn it.

Read more →
Reflection

Rules control, not roles.

Why access permissions create bottlenecks — and what works instead.

Read more →
Technical

How I reduced AI context from 30K to 3.5K tokens.

The trick isn't better summarization — it's compilation.

Read more →
Technical

A democratic system for AI memory maintenance.

No manual cleanup. No curator. The system maintains itself through consensus.

Read more →
Technical

How I tested an AI system with 4 different models.

52 tests each, no human intervention. Here's what happened.

Read more →
Tutorial

From zero to persistent AI memory: set up EIDARA in 10 minutes.

No technical skills required. Download, hand it to any AI, done.

Read more →
Tutorial

How to share AI agents across your team.

An agent is a .md file. Share it like any document.

Read more →