Simon

mts @reflection_ai working on mirror

[Simon @depthfirstsearc]

@kwindla Congrats team!

[Simon @depthfirstsearc]

@nic_carter The vickrey auction is one of the greatest game theory inventions of all time. So sad that they’ve silently done away with it

[Simon @depthfirstsearc]

@camworboys @Block this is the most impressive demo I've ever seen in voice. Congratulations

[Simon @depthfirstsearc]

@RagavanThurair wow

[Simon @depthfirstsearc]

@chaotictransfem how can i do this

[Simon @depthfirstsearc]

@badlogicgames yes.

[Simon @depthfirstsearc]

@TheStalwart Is this not always the case with new technology?

[Simon @depthfirstsearc]

@kylascan This is so amazing.

[Simon @depthfirstsearc]

@uncledoomer connecticut runs the world

[Simon @depthfirstsearc]

@nicochristie @BrainsAndTennis awesome stuff

[Simon @depthfirstsearc]

@eva_roytburg @citrini Congrats!!

[Simon @depthfirstsearc]

@messyfork Sorry I'm doing something that thursday

[Simon @depthfirstsearc]

The ai slowdown has already been implemented and it’s called the gta 6 release

[Simon @depthfirstsearc]

@ctatedev lol what

[Simon @depthfirstsearc]

@dexhorthy +1

[Jake Nations @nayshins]

Mirror is a Rust agent core with a python TUI using an RPC. The same agent core can run headless behind an HTTP API just as easily as in your terminal.

Every session is a single append only log, and everything else (the transcript, what the model actually sees, token usage, extension state) is just a fold over that log.

Compaction appends a new context window with a summary and the recent turns, and a rewind is a fork that points back at the parent’s log up to where you rewound, so nothing ever gets copied or thrown away.

[Simon @depthfirstsearc]

More on how mirror works from the creator himself @nayshins !

[Jake Nations @nayshins]

Mirror is a Rust agent core with a python TUI using an RPC. The same agent core can run headless behind an HTTP API just as easily as in your terminal.

Every session is a single append only log, and everything else (the transcript, what the model actually sees, token usage, extension state) is just a fold over that log.

Compaction appends a new context window with a summary and the recent turns, and a rewind is a fork that points back at the parent’s log up to where you rewound, so nothing ever gets copied or thrown away.

[Jake Nations @nayshins]

Ok can finally talk about what I’ve been working on at Reflection. Introducing Mirror, our own open source coding agent optimized for Beam.

I started the mirror project when I first joined Reflection, and it has gone through many iterations since then.

Install it today with: uv tool install reflection-mirror
https://t.co/hhPuEML1bX

[Simon @depthfirstsearc]

@rauchg @vercel @eve @fx lfg!

[Simon @depthfirstsearc]

Side note: this is such a good graph

[Artificial Analysis @ArtificialAnlys]

Beam scores 34% of on the Artificial Analysis Cyber Index, making it the leading US open weights model that we’ve tested

Driven by its token efficiency, it also sits on the Cyber Index vs. Output Tokens per task frontier, alongside GPT-6 Sol (Daybreak Blue)