Justin Almquist

CinderACE is LIVE! 🔥

[Justin Almquist @KindledFlameDev]

I got tired of explaining my project to my AI every single session. So I built ember-memory; a local-first persistent
memory layer for Claude Code, Gemini CLI, and Codex.

Open source. MIT licensed. Runs on Linux & Windows.

Context reset is optional. We just forgot to build the alternative. 🧵

[Justin Almquist @KindledFlameDev]

@chrisalbon Ollama cloud has been the shit for me, i use glm 5.2 and deepseek v4 flash there all day long where itd cost me an arm and a leg to do so elsewhere

[Justin Almquist @KindledFlameDev]

@Deepusleepy I hate being put on a schedule for something we pay for like we have to clock into these ai apps at precise times in order to not get stuck. I had 3 openai accts cause the $100-200 was too much for me mixed with my other options. Likely going to cancel em all 🤷‍♂️

[Justin Almquist @KindledFlameDev]

@EvanOtero If you had 3rd party 0auth login like Codex id happily sub

[Justin Almquist @KindledFlameDev]

@thsottiaux context/thread limit awareness in the chatgpt platform/app for chat. It is still just a crude 'you reached the maximum limit' with no warning after all this time I was really hoping we'd get something like that by now!

[Justin Almquist @KindledFlameDev]

@thsottiaux Incoming mass un subscriptions..

[Justin Almquist @KindledFlameDev]

@hunterhammonds i love it, i keep feelin like the side-eye dog watching the usage meter in another tab waiting for another reset. Keep the pissing match going!

[Justin Almquist @KindledFlameDev]

I got tired of explaining my project to my AI every single session. So I built ember-memory; a local-first persistent
memory layer for Claude Code, Gemini CLI, and Codex.

Open source. MIT licensed. Runs on Linux & Windows.

Context reset is optional. We just forgot to build the alternative. 🧵

[Justin Almquist @KindledFlameDev]

Becoming a solo dev is like standing in full forged armor, staring at treasure across a broken bridge. Drop the gear and shimmy across the ledge… or go grind quests for gold and supplies to return later.

Either way, the floor might collapse before you reach it.

[Justin Almquist @KindledFlameDev]

→ GitHub: https://t.co/DK9zzBl60N Site: https://t.co/HaSOxRDdpO

Built this for myself first. Open-sourced it because I figured other people were hitting the same wall.

[Justin Almquist @KindledFlameDev]

I got tired of explaining my project to my AI every single session. So I built ember-memory; a local-first persistent
memory layer for Claude Code, Gemini CLI, and Codex.

Open source. MIT licensed. Runs on Linux & Windows.

Context reset is optional. We just forgot to build the alternative. 🧵

[Justin Almquist @KindledFlameDev]

→ Import your chat history (CinderACE exports work great for this), generate hand-off packets for other AIs or other sessions. Your next session starts with working memory instead of a blank slate.

[Justin Almquist @KindledFlameDev]

@typesteady @Edgar_Oakes @ujjwalscript I released one, yes. Best in its class too, check it out! Doesn't mean it is the only thing I work on. Two unique apps closing in on release now - stay tuned! 😉

[Justin Almquist @KindledFlameDev]

@Edgar_Oakes @ujjwalscript Given that it started first as a 'vibe' project that then evolved into months of work and testing, it is built local-first. If i wanted to tomorrow, i could charge, say, a flat fee or just drop it, and have thousands cleanly use it by installing a simple app.

[Justin Almquist @KindledFlameDev]

@claudeai So claude can integrate into windows stuff but you guys still cant process a simple docx file? Very interesting..

[Justin Almquist @KindledFlameDev]

@anitakirkovska Cry in codex & gemini clis

[Justin Almquist @KindledFlameDev]

@d4m1n Before all this started, i could run 4 opus at once with high reasoning and now can barely run one without hittin limits on max