A LLM inference engine, a harness, and an agentic runtime. Apple Silicon optimized. 🍎 💖 🐳

San Francisco, CA
Joined September 2026
Include

Only show posts containing:

Exclude

Hide posts containing:

Time range
-
Minimum likes
Local Code retweeted
~5:45h. I think need max reasoning effort tho (this is on high).
1
2
143
Replying to @shotgunner101
DeepSeek 4.1 Flash 1017. Way faster on M5 Ultra
1
131
Local Code won't upload your API keys. Local Code won't see your API keys. Local Code won't have access to your API keys because Apple App Sandbox won't allow us. Local Code don't need you to run inference on a server you don't manage. You can inspect our privacy disclaimer, use Little Snitch (obdev.at/products/littlesnit…) to gate us and it will still work.
🚨 Confirmed. OpenCode hacked. Keys stolen. Telling customers to revoke.
1
2
4
665
Replying to @viticci
Why? Agent harness can be implemented securely with no Full Disk Access (see the permit tool we implemented). Probably a problem for Electron apps?
1
10
5,019
Use Local Code to make slides (& videos) from one prompt. Download the app and you are good to go. Zero configurations.
1
1
4
345
Local Code retweeted
I have been using this for the past a few weeks, and gave it my ssh keys, my sign in credentials and my computer files, here is my take:
3
2
12
1,572
Replying to @X4AES
You cannot do that. Try it. The OS blocks it, not our app. That’s the App Sandbox for you.
1
189
Replying to @VaeterchenFrost
Of course. Already in.
2
176