Did a deep dive into why Blueprints are actually being deprecated in #UE6: tobosoksini.games/blueprints-w...
#UnrealEngine5 #UnrealEngine6 #GameDev #Blueprints #Verse #GameEngine #UE5 #AI #AIagents #CodingAgents

Did a deep dive into why Blueprints are actually being deprecated in #UE6: tobosoksini.games/blueprints-w...
#UnrealEngine5 #UnrealEngine6 #GameDev #Blueprints #Verse #GameEngine #UE5 #AI #AIagents #CodingAgents
My current work setup feels kinda weird: I put AI to work on all my Macs sitting at home.
All I take with me is an iPad to give it tasks.
Had to get out of the house so I wouldn’t distract it.
Have a great Sunday, everyone!
Eight flaws across seven CLI coding agents: a repo's own .git config names a command the agent runs on your machine, outside the sandbox, before the trust prompt. Only bites when the repo arrives as files, not a clone. Read the filing #git #codingagents #supplychain #infosec
One index. Many writers. Delayed merges through Git.
Two unrelated changes can conflict simply because they land at the same place in the file.
My fix in Keep the Why: deterministic A-Z / 0-9 write areas.
Simple idea, surprisingly useful.
This time, I was able to create this cute little low-poly section of a metropolis, in one-shot, using a tiny model: qwen-3.8-27B with flash-3D.
More demos of flash-3D: bsky.app/profile/frea...
#AI #3D #CodingAgents #GenerativeAI #lowpoly #3D #astra #blender #flash3d
It’s trying to run the entire engineering loop.
And because this one comes from Microsoft, this is definitely a project worth watching.
#Microsoft #AIAgents #AI #CodingAgents #Codex #ClaudeCode #SoftwareEngineering #OpenSource #DeveloperTools
Already using Kong? Keep the existing API edge and evaluate whether the LLM layer needs specialized routing, guardrails and budget controls.
Read the nRouter comparison:
A router recommends. A gateway executes.
See how NotDiamond’s model-selection layer differs from nRouter’s managed request path for provider access, budgets, guardrails and logs.
Keep a Changelog tells you what changed. Keep the Why keeps the reasoning behind it: decisions, rejected ideas, constraints and workarounds, stored as Markdown in the repo.
Project memory for humans and coding agents.
No database. No daemon. No account.
Your AI stack doesn’t need separate infrastructure for every provider.
nRouter brings model access, smart routing, fallbacks, guardrails, observability, and budget controls into one OpenAI-compatible gateway.
Explore: nrouter.ai
Specialist observability or a unified LLM gateway?
Langfuse focuses on observability, prompts, and evaluations. nRouter combines observability with routing and governance.
Watch today's Century Report podcast here:
OpenAI's coding agents now log 3.1 workdays for every human workday, and its chief scientist calls the systems an alien mind. #CodingAgents #AIBenchmarks https://sharedsapience.com/century-report/the-century-report-september-7-2026/
An LLM experiment isn’t deterministic if the same user keeps switching variants.
Hash a stable identifier to preserve assignment across calls and produce cleaner comparisons.
Read the guide: nrouter.ai/blog/enginee...
The most token-efficient 3D-modeller for coding agents. Everything shown here was one-shotted using Flash-3D. No blender needed. Open-source.
#AI #3D #CodingAgents #GenerativeAI #lowpoly #3D #astra #blender #flash3d
Don’t route based on headline model prices.
Route by task, test on your own workload, and compare measured cost per successful output against proven quality.
Read the guide: nrouter.ai/blog/guides/...
Can your LLM gateway back its security claims with evidence? Start with this 6-area checklist. nRouter’s SOC 2 Type II audit is in progress.