THE 65 CONNECTION retweeted
Only on RollingOut.com/ReasonableDou…
Black women deserve the space to be powerful, vulnerable, brilliant and fully human. 🖤
Our latest cover star, Emayatzy Corinealdi, brings that complexity to the screen as Jax Stewart in Reasonable Doubt, commanding attention even when her character’s world begins to unravel.
In her Rolling Out cover interview, Corinealdi opens up about the emotional cost of telling Jax’s story, confronting addiction with compassion, and allowing Black women characters the freedom to fall, fight and find their way.
The presence is undeniable. The conversation goes deeper.
Can you root for Jax while holding her accountable? Let’s talk. 👇🏾
#RollingOut #EmayatzyCorinealdi #ReasonableDoubt #JaxStewart #BlackWomenInHollywood CoverStory
THE 65 CONNECTION retweeted
Released on November 12th, 1980 by @warnerbros.
Ultra Wave is the fifth studio album by Bootsy Collins. Three singles were released from this project.
Mug Push/Scenery (Warner Bros. WBS 49599).
F-Encounter/Instrumental with Rap) (Warner Bros. WBS 49661).
Produced by George Clinton and William Collins.
12 AI Prompts for Clothing Brands 👇🏻 pin.it/6kcb91RVy via @pinterest
69
Grok Bot could be the operating layer AI agents have been missing.
It can use tools, run routines and coordinate other agents with minimal supervision.
In this article, I show you how.
Elon Musk (SpaceXAI founder) released a prompt that turns one Grok Bot into an entire project team
It's f*cking unreal...
Bookmark and send it to your Grok Bot, fill in the project brief and thank me later
Then read the full roadmap with examples below
THE 65 CONNECTION retweeted
My 2026 AI Engineer learning stack
Python → fundamentals
FastAPI → APIs
PostgreSQL → data
Redis → caching
LLMs → intelligence
RAG → knowledge
Vector search → retrieval
Tool calling → actions
MCP → integrations
Evals → reliability
Docker → deployment
Observability → production
Don't learn 40 frameworks.
Learn how the pieces fit together.
THE 65 CONNECTION retweeted
MCP, RAG, and AI Agents solve different problems:
MCP → connects AI to tools & systems
RAG → connects AI to relevant knowledge
AI Agents → connect reasoning, tools & actions
Different layers. Different jobs.
Together, they form the foundation for building useful AI systems.
THE 65 CONNECTION retweeted
10 GITHUB REPOS FOR AI VIDEO GENERATION:
1. ComfyUI
github.com/comfyanonymous/Co…
2. Diffusers
github.com/huggingface/diffu…
3. Open-Sora
github.com/hpcaitech/Open-So…
4. Wan2.1
github.com/Wan-Video/Wan2.1
5. HunyuanVideo
github.com/Tencent-Hunyuan/H…
6. CogVideoX
github.com/THUDM/CogVideo
7. LTX-Video
github.com/Lightricks/LTX-Vi…
8. Mochi
github.com/genmoai/models
9. VideoCrafter
github.com/AILab-CVC/VideoCr…
10. AnimateDiff
github.com/guoyww/AnimateDif…
SAVE THIS STACK
THE 65 CONNECTION retweeted
whoever dropped this is f*cking crazy
someone organised 300+ AI agents, frameworks and tools into one GitHub repo.
coding, browser automation, memory, voice, research, multi-agent systems, local models, sandboxes and evals.
one place to look before you spend a weekend rebuilding something that already exists.
the categories are where it gets useful.
need help working inside an existing codebase?
→ Aider / Continue
need several agents handling different parts of a task?
→ CrewAI / AutoGen / MetaGPT
need memory that carries across conversations?
→ Mem0
need a sandbox for executing agent-generated code?
→ E2B
want to run models locally without a hosted inference bill?
→ Ollama
need to trace what happened when an agent goes off-script?
→ AgentOps
and those are just a few entries.
the list spans roughly 25 categories, so you can look for the specific piece your stack is missing instead of scrolling through another random tool dump.
don’t install 300 projects.
find the one that saves you from building the wrong thing.
check whether it’s maintained, read the limitations and try it on one real task.
a repo being listed doesn’t mean it’s production-ready.
but knowing it exists can save you a lot of wasted work.
bookmark the map before your next build.
your first question should be:
“what can I reuse?”
not:
“how do I build all of this from zero?”
THE 65 CONNECTION retweeted
holy sh*t, a SpaceXAI engineer shared his personal setup for running 200+ coding agents
Lingxi Li says 5 specialist Grok Bots manage the fleet
each owns an area of the codebase and launches Cursor cloud agents to do the work
the part I’d steal is the follow-up loop:
01 give every bot an area
mobile / desktop / infrastructure / Android / the harness
clear ownership + focused context
02 make it ask for proof
screenshots, transcripts and a working dev instance
the bot checks the result and sends corrections back
03 keep the work on a board
every 30 minutes → check CI, conflicts and review findings
unfinished work goes back to the agent
04 give mistakes an owner
his ops bot, Jenny, runs postmortems and updates the shared playbook
my smaller version: one owner, one coding agent, one real bug
paste this brief into your engineering bot ↓
Own [AREA] of [REPO]. Reproduce [BUG], then launch a cloud agent using the relevant repo skills. Require before/after evidence for [USER FLOW] and a PR. Inspect the result, follow up on failures, and keep the task board current. Send me the PR and checks for review; leave merging to me
save this, then build a harness for your Grok bot ↓
THE 65 CONNECTION retweeted
this might be the most useful AI agent paper of 2026, and it's certified f*cking gold
Wavestone AI Lab took apart Claude Code, Codex and 9 more agents
every good one runs on the same 7-part harness
agent = model + harness
all 11 use the same 7 parts. only the size changes
the 7 parts:
> loop - from a plain while loop to fully replayable runs
> LLM layer - one prompt template up to 29 provider profiles
> tools - bash only up to 43 typed tools on demand
> memory - full history up to notes kept across sessions
> safety - a step limit up to rules + reviewer model + sandbox
> orchestration - Aider skips sub-agents, others fork them
> extensions - SKILL.md in 9 of 11, MCP in 8
it ends with a 90-line harness covering all 7
the cleanest starting point for your own agent
the pages worth your time:
> p.4 - the 7-part map
> p.12 - 3 loop types, Claude Code and Codex as examples
> p.53 - why nobody uses frameworks or embeddings
> p.67 - 18 design rules
> p.71 - the 90-line harness to copy
Andrej Karpathy predicted the future of AI once again:
“Everyone’s renting frontier models for jobs a local 3B model could do. Small Local LLM's are the future.”
this 18-page PDF breaks down Karpathy's case for running that core on your own machine.
the real question isn't "is the local model as smart?"
It's "which of my 1000 calls ever needed the smart one?"




































