Usman Pervaiz

AI Agent Developer 🤖 | MERN Stack Dev (4+ yrs) 👨‍💻 | OpenAI, n8n, ElevenLabs, LangChain, Pinecone,Claude ,Vector DBs , openclaw

[Usman Pervaiz @usmanpervaizdev]

￣￣￣￣￣￣￣￣￣￣￣￣￣|
| Developers Let's Connect! |
|＿＿＿＿＿＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
——
| |
|_ |_

#100DaysOfCode

[Gaurav Sen @gkcs_]

How to train a transformer.

I am a software engineer who has no prior experience with ML experience.

But now I have trained multiple nueral nets for different types of applications.

If you are looking to learn ML or AI, this is the best time.

https://t.co/RcPgrhVTDz

[Sergii Makarevych @sermakarevich]

Agent guardrails cheat sheet

[Akshay 🚀 @akshay_pachaar]

Google just open-sourced Harness².

(turn any harness into a self-improving system)

a harness is everything wrapped around the model, like the system prompt, tools, memory, skills, and subagents. today a person designs it once, and the agent runs every new task with that same fixed setup.

Harness² lets the agent refine that setup for each new task on its own, with the model weights left untouched.

on every task, the agent proposes edits to its harness, runs a few candidate versions, and composes a better harness by comparing how those runs behaved.

because it learns from the differences between runs, it needs no correct answer or grader for the task.

that covers a single task. Harness² also carries what it learns from one task into the next.

after each task, it saves the harness it settled on along with the lessons that held up across the runs. the next time it refines a harness, it starts from those notes instead of from scratch.

this means every finished task makes the next round of edits more informed.

Google tested this with Gemini 3.5 Flash and GLM-5 on legal, office, and professional work benchmarks. one round of edits was enough to beat their original harnesses, and giving the agent more rounds pushed scores higher.

those extra rounds can go wide or deep. the agent either tries several candidate harnesses at once or keeps refining a single one step by step, using OpenCode or Codex to carry out the work.

you can find the Harness² repo here: https://t.co/i00WLeeFpw

I wrote about a similar idea for agents already running in production, where the harness should repair itself after a failure instead of waiting for someone to debug the trace by hand.

that piece walks through how @Cometml's Opik (open-source) closes this loop, going from a failing trace to a proposed fix to a regression test that stops the same failure from coming back. it's quoted below.

[Usman Pervaiz @usmanpervaizdev]

𝙏𝙝𝙚 𝙠𝙚𝙮 𝙡𝙚𝙨𝙨𝙤𝙣:
A longer context window does not guarantee better understanding. The real challenge is making models use every part of the context effectively.

𝗣𝗮𝗽𝗲𝗿:
https://t.co/Lz7YJFsuBU

[Usman Pervaiz @usmanpervaizdev]

𝗖𝗮𝗻 𝗹𝗮𝗿𝗴𝗲 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗺𝗼𝗱𝗲𝗹𝘀 𝗿𝗲𝗹𝗶𝗮𝗯𝗹𝘆 𝘂𝘀𝗲 𝗶𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗶𝗻𝘀𝗶𝗱𝗲 𝗹𝗼𝗻𝗴 𝗽𝗿𝗼𝗺𝗽𝘁𝘀?

[Usman Pervaiz @usmanpervaizdev]

3. Increasing the context window size does not automatically mean better reasoning.   

4. More retrieved information can sometimes reduce accuracy instead of improving it.   

5. The position of information matters as much as the amount of information provided.

[Usman Pervaiz @usmanpervaizdev]

Want to explore how agent harnesses work internally?

Check out these open-source projects:

Pi: https://t.co/UNYcxvtdu5

OpenHarness: https://t.co/v0gDPftWcl

[Usman Pervaiz @usmanpervaizdev]

Over the last 3 years, I have mostly worked with LLM wrappers.

One thing I have learned: building an AI agent is more than connecting an LLM to tools.

The key is the agent harness.

[Usman Pervaiz @usmanpervaizdev]

Each component plays a role in making agents useful and reliable.

The harness manages tool calls, provides context, controls access, and handles execution.

[Andrej Karpathy @karpathy]

Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.

- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.

The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.

So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.

[Usman Pervaiz @usmanpervaizdev]

Most of the time we let the LLM to summarize when compacting the context . but in a harness , Pi and Hermes write a fixed-structure markdown file instead.

[Usman Pervaiz @usmanpervaizdev]

𝗕𝗶𝗴𝗴𝗲𝗿 𝘄𝗶𝗻𝗱𝗼𝘄𝘀 𝗱𝗼 𝗻𝗼𝘁 𝗳𝗶𝘅 𝗰𝗼𝗱𝗶𝗻𝗴 𝗮𝗴𝗲𝗻𝘁𝘀. 𝗧𝗵𝗲 𝗵𝗮𝗿𝗻𝗲𝘀𝘀 𝗱𝗼𝗲𝘀.

1. When does compaction start?
2. Does compact erase history?
3. Context rules vs isolation?

https://t.co/P3ScO2CIuO

#AI #LLM #AIAgents #ContextWindow

[Andrew Ng @AndrewYNg]

New: A map of the most important skills in AI Engineering. https://t.co/VVkn1Dqp1N

[Usman Pervaiz @usmanpervaizdev]

Recently I have been working on one of my projects
and while doing that I looked into how Pi handles security.

The operation layer makes it more secure by controlling the switching of tools between SSH and the sandbox.

[Usman Pervaiz @usmanpervaizdev]

If you write JavaScript, learn how it actually executes.

Call Stack → Heap → Web APIs → Event Loop

[H’s mom @aiishadahir]

It’s never luck, it’s always Allah