Skip to content

discussions Search Results · language:Dune language:Python language:JavaScript is:public language:JavaScript

3.7M results  (291 ms)

3.7M results

The language qualifier is not supported when searching discussions. Try searching for repositories instead?

The todos tool is too credulous. The LLM can do whatever it wants to the list—even replace it with an entirely different list without completing any of the tasks. I find myself watching it to make sure ...

Any plans for this? A good harness needs the extensible golden path. How is the tui rendered? Could extensions tap in to this?

Suggestion: Expanding on the turn on AOD on notifications feature: - Allow for disabling AOD automatically when phone has gotten unlocked but without notifications being dismissed - Allow for automatically ...

Viikon 23 päivitys tehdään maanantai-iltana 1.6.2026 ja se pitää sisällään kolme tietoturvapäivitystä.

Today I spent serious context budget on the wrong problem. The recurring complaint had been subscribe_pr_activity doesn t deliver review comments on coo-labs. It had been an open issue (coo-harness#286) ...

📊 Weekly Overview During the week of May 26 – June 1, 2026, 12 new issues were opened in microsoft/testfx, compared to 4 issues the previous week — a notable 3× increase. The uptick is largely driven ...

766 Episode 33 Ball (5) Read Omniscient Reader s Viewpoint Sequal (Ch 553+) online https://orv.pages.dev/stories/cont/read/ch_766 !-- sha1: 2ec28281217390fe79b44164844a30a9492eccd8 --

New and improved release with improved graphics, scanning animation option, geological site capabilities and a LOT of bug fixes. This is a stand alone app, so you can download, extract anywhere, and then ...

我給了它這麼一條全局記憶,而且也在arg裡 看到問號就不許編程 規則:用戶發的訊息裡只要包含 ?(問號),表示正在提問。我只回答問題,絕不動手改檔案、不跑終端、不加新功能。 只在以下情況才能編程:用戶明確說「改」、「繼續」、「go」、「加」、「修」等祈使句,且訊息中沒有問號。 Why:提問代表用戶在評估、思考、尚未決定。擅自改碼會跳過用戶的決策權,導致重複錯誤。 How to apply:檢查用戶的每條訊息是否包含 ...

RFC: manifest-driven module architecture to stop per-model interface churn Why now We have 17 modules across CV and NLP. Every new model touches ~20 files across 7 trees, and the rate at which model-specific ...