DEV Community

#dailybuild2026

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
Your agent is a long-running process with side effects. Kill it and watch what happens.

Your agent is a long-running process with side effects. Kill it and watch what happens.

Comments
11 min read
Sleeptime memory consolidation: an agent that edits its own memory and proves it still remembers

Sleeptime memory consolidation: an agent that edits its own memory and proves it still remembers

Comments
15 min read
Luthier: an agent that audits an AI coding agent's harness against the repo it governs.

Luthier: an agent that audits an AI coding agent's harness against the repo it governs.

Comments
9 min read
One UI, Any Agent: Proving Framework Independence with AG-UI, Mastra, and React

One UI, Any Agent: Proving Framework Independence with AG-UI, Mastra, and React

Comments
5 min read
Bloom: I Made Two LLMs Paint the Same Sentence and Measured What Happened

Bloom: I Made Two LLMs Paint the Same Sentence and Measured What Happened

Comments
7 min read
Building Clay: an A2UI host where the model emits interface instead of prose

Building Clay: an A2UI host where the model emits interface instead of prose

1
Comments 1
11 min read
Glyph: I made two AI models fight inside real terminals and the only honest judge is a PTY

Glyph: I made two AI models fight inside real terminals and the only honest judge is a PTY

1
Comments
4 min read
Building The Confidence Curve

Building The Confidence Curve

1
Comments
14 min read
Automating LLM A/B Testing: How to Build a Cross-Model Evaluation Harness Programmatically

Automating LLM A/B Testing: How to Build a Cross-Model Evaluation Harness Programmatically

Comments
6 min read
How I built Refract: a live arena where two LLMs fight in GLSL

How I built Refract: a live arena where two LLMs fight in GLSL

1
Comments
4 min read
I measured how deterministic LLMs actually are, prompt by prompt. The results broke my hypothesis.

I measured how deterministic LLMs actually are, prompt by prompt. The results broke my hypothesis.

1
Comments
10 min read
Two Models, One Board: Building "Model vs Model"

Two Models, One Board: Building "Model vs Model"

1
Comments
16 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.