Saransh

AI|ML

[Shubham Saboo @Saboo_Shubham_]

wtf is a dynamic agent orgs.

Self-evolving agent orgs where graph rewrites itself while the work is happening.

[Saransh @SaranshSaboo]

@Saboo_Shubham_ Great share 🔥

[Shubham Saboo @Saboo_Shubham_]

https://t.co/xjGmEGMxtW

[Shubham Saboo @Saboo_Shubham_]

This is HOW we build, test, and scale Agent Skills at Google.

Open-source Google Agent skills GitHub Repo: https://t.co/fbv91QBcA3

[Shubham Saboo @Saboo_Shubham_]

Someone just ran Kimi k3 with 2.78 trillion parameter on a single CPU in 8.24 GB of RAM. 100% Opensource.

[Saransh @SaranshSaboo]

@Saboo_Shubham_ @KairosDataLabs Thanks for sharing!

[Shubham Saboo @Saboo_Shubham_]

Watch the full Agent Zero Episode on YouTube.

https://t.co/DsHS1psglk

[Shubham Saboo @Saboo_Shubham_]

Loop Engineering clearly explained in just 30 minutes.

I broke down everything I know about building loops that run without a human watching.

What every real Loop needs, loops you can copy today, and why loops break in production.

Available now for FREE on YouTube.

[Shubham Saboo @Saboo_Shubham_]

MCP just went stateless. And it makes so much sense.

[Shubham Saboo @Saboo_Shubham_]

WILD times.

Anthropic: Opus 5 beats GPT-5.6 on ARC-AGI-3
Tibo: I change two settings and GPT-5.6 Sol is now SoTA

[Shubham Saboo @Saboo_Shubham_]

Google just shipped Open Knowledge Format v0.2, and it is all about agentic trust.

When a person writes a wiki page, you know who to ask if it is wrong. When an agent writes ten thousand pages overnight, nobody is accountable for any of them. The agent reading those pages needs signals it can check before it reads a single line.

v0.2 puts four of them in the frontmatter:

• sources: what the page was built from, with author and usage count. Raw signals, no trust score.
• generated and verified: who wrote it, who checked it. Human checked and machine checked stay separate.
• stale_after: a fixed date, so freshness is a date comparison.
• status: draft, stable, or deprecated.

The best part is a new type called Attested Computation. It holds the approved query for a metric. The agent fills in parameters and nothing else. It never writes the SQL. After a run, a small script compares what actually ran against what was approved. Rename a table or drop a join and it fails, and the number never gets shown.

Still just markdown and YAML. Every new field is optional. Your v0.1 bundles still work.

GitHub Repo: https://t.co/BlwunQhi7q

[Shubham Saboo @Saboo_Shubham_]

Andrew Ng's open-source AI coworker is the WAY.

Works with GPT, Claude, Gemini, GLM and even the local models using Ollama.

Comes with 25+ integrations and you can use it directly from slack.

100% Opensource.

[Saransh @SaranshSaboo]

@Saboo_Shubham_ @unwind_ai_ Great share!

[Shubham Saboo @Saboo_Shubham_]

INSANE. This gets better ever other week.

Opus 5 as advisor, GPT-5.6 as orchestrator, and Gemini 3.6 Flash is the worker.

Let that sink in.

[Shubham Saboo @Saboo_Shubham_]

Every developer needs this new Agent Skill.

Say what you meant to fix. It flags everything else in your GitHub diff.

Runs locally. 100% Open-source. Super cool SHIP by @mvanhorn.

[Shubham Saboo @Saboo_Shubham_]

AISLOPOPEDIA

Created this sometime ago just for fun.

https://t.co/BqMcyz7oIQ

[Shubham Saboo @Saboo_Shubham_]

This is now an open-source Agent skill.

Thinking out Loud.

Ramble at your agent. Audit what it heard before it acts.

[Shubham Saboo @Saboo_Shubham_]

Excalidraw for AI Agents....INSANE.

[Saransh @SaranshSaboo]

@so_ainsight awesome-llm-apps 🔥

[Saransh @SaranshSaboo]

@Saboo_Shubham_ Thanks for sharing!