Every reply in a long ๐๐ฒ๐ฟ๐บ๐ฒ๐ Agent chat sends the entire chat to the model again. ๐คฏ
This is a vibe coding post classified by AGK Intelligence as Agent infra & costs (a workflow), kept by the Vibe Coding Radar because it carries real work, not commentary.
Every reply in a long ๐๐ฒ๐ฟ๐บ๐ฒ๐ Agent chat sends the entire chat to the model again. ๐คฏ That's why reply 200 feels slower than reply 2. These 5 settings keep it short. 5 ๐๐ฒ๐ฟ๐บ๐ฒ๐ settings for when long chats get slow๐ 1. Finished one part of the job? โ /compress here 2 โธ Summarizes everything before your last 2 exchanges โธ Keeps those 2 word for word, so the thread survives Needs: nothing. Works in the terminal and in Telegram. 2. Big 1M-window model? โ compression.threshold_tokens 256000 โธ By default it waits until ~500K tokens to compact โธ This compacts at 256K, so each reply re-reads less Nee
Posted by Hermes Super-Intel (5.1k followers) 3 days ago ยท 49 likes ยท 3.9k views ยท view the original post on X. Kept by the Vibe Coding Radar as Agent infra & costs.
More vibe coding work like this
- The tricky part of sandboxing is often what happens at the boundary with the host. โ @AgenticAIFdn
- 575 - Eight or nine repos running at once, and one agent kept dragging memories from theโฆ โ @PowerBITips
- If you want to learn how to build self-improving harnesses, check out this paper. โ @dair_ai
- Build an agent on @onkeiki to answer your customers while you sleep. โ @orchid_hq
- how do you stop 20+ agents from colliding on the same client work? โ @webjuice_ie
- Brilliant paper on optimizing agent environments. โ @dair_ai
- Docker is one box. The coding agent can run in others, and switching is one line ๐งฐ โ @marcelotryle
- Holy f*cking sh*t... everyone upgrades Claude and nobody builds the harness โ @gippp69
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 7.9k posts from 5k X accounts over the last 14 days, 2.6k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour โ last update 2026-11-06 16:53 UTC. Full method.