Muratcan Koylan

MTS, AI Research @sullyai

Building AI medical workforce

Prev/ AI Persona Development, HCI, Marketing, Communication Design

[Muratcan Koylan @muratcan]

I’m excited to share a new repo: Agent Skills for Context Engineering

Instead of just offering a library of black-box tools, it acts as a "Meta-Agent" knowledge base. It provides a standard set of skills, written in markdown and code, that you can feed to an agent so it understands how to manage its own cognitive resources.

https://t.co/vWwrYPAd8k

Most agent failures are not model failures; they are context failures. This is still an experimental project. The goal is to establish a platform-agnostic standard for context engineering that can be used in Cursor, Claude Code, Copilot or Codex.

skills/
context-fundamentals: What context is, why it matters
context-degradation: How context fails (lost-in-middle, poisoning)
context-optimization: Compaction, masking, caching
multi-agent-patterns: Orchestrator, swarm, hierarchical
memory-systems: Vector RAG, knowledge graphs, Zep
tool-design: Building tools agents can use
evaluation: Testing and measuring agent systems

I believe this is a good start, showing developers how to approach context engineering rather than relying on ready-made tools.

You will also find the aggregated research documents I used to build these skills in the repo. The skills are synthesized from technical blogs on context and prompt engineering that I bookmarked, AI Labs' documentations, and Anthropic Skills examples.

Try the 7 Skills, created using Antrhopic's Skills template format. Experiment with the provided scripts and references, and feel free to contribute to the repo.

[Muratcan Koylan @muratcan]

It’s actually a good question; the difference is subtle but structural.

I usually frame it like this:
AGENTS[.]md acts as the declarative context. You write this for every repo (and nested directories) to define the project structure, persona, and coding rules.

Skills are the functional protocols. They provide the agent with modular capabilities like advanced tool-use and multi-step chaining that are dynamically discovered only when needed.

If AGENTS[.]md defines the identity and environment (the body), Skills provide the specialized toolset (the capabilities) used to execute tasks autonomously.

[Muratcan Koylan @muratcan]

@rauchg Same here. Most of the token cost goes toward open-source inference.

[Muratcan Koylan @muratcan]

@atakoylanai Good post cuz

[Muratcan Koylan @muratcan]

Ata's hands-on experience on film sets, now shaping how he makes AI ads and movies.

Go follow my cousin if you’re working in this space. He shares the creative decisions behind the work.

He’s also here with me in Toronto for a while. If you’re around and want to collaborate, hit him up.

And if you can help with video model credits, we’ve got plenty of ideas!

[Ata Koylan @atakoylanai]

My AI films got better when I stopped turning every generated image into a video.

Building a scene > animating a still

Here's how I created this ad using Seedance 2.5 and Nano Banana Pro, from character design to final edit:

[Muratcan Koylan @muratcan]

@victorialslocum @weaviate_io Congrats Victoria!

[Muratcan Koylan @muratcan]

@kevin_t_ngo Great run, keep going

[Muratcan Koylan @muratcan]

@CompleteSkeptic @HamelHusain How does it perform as a background agent in a duplex voice architecture?

[Muratcan Koylan @muratcan]

@MarkJCarney Let’s build in Canada 🫡

[Muratcan Koylan @muratcan]

@thejamescad @profound @sequoia @kleinerperkins Wow, Profound has really come a long way, congrats on the launch!

[Muratcan Koylan @muratcan]

@thaiscbranco_ Congrats on the launch!

[Muratcan Koylan @muratcan]

@natalieyeo @OpenAIDevs Not this year unfortunately, visa issues 🥲

[Muratcan Koylan @muratcan]

@arizeai using an ai slop generator bot as an ai company is not a good look tbh

[Muratcan Koylan @muratcan]

When I tried GPT-Live-1 today, I felt deflated as I recognized behaviors we’d spent weeks trying to make reliable. For the past three months, I’ve been focusing on researching and building duplex voice agents, where I wrote thousands of lines across our harness and the experiments around it.

Customers use parts of that work every day, the conversions are increasing drastically as we make our voice agents more reliable and expressive. This duplex voice harness/model (interaction runtime + kernel) problem has occupied most of my attention but tbh when I first started working with voice agents, I didn't know how challenging and complex they were. Duplex conversational timing, orchestration, acoustic turn-taking, micro-interruptions, prosody, and generalized noise etc are really hard problems.

There are so many 'beautifully crafted"'voice demos but seeing them in production is almost impossible. A cough would interrupt the voice agent, it leaves the caller waiting for it to speak again. During a tool call, someone would change their decision before the earlier result returned. There are also many other telephony factors so it is not just an AI or research problem.

Following those cases through the system led us to separate the part that speaks from the work running in the background. We’ve developed a kernel that controls what the agent can do and checks what happened before it reports an outcome to the caller because the duplex models simply weren’t good enough yet.

During this period, reading NVIDIA’s PersonaPlex and Nemotron VoiceChat work helped me understand where the models were heading; Thinking Machines’ Interaction Models approach to keeping an interaction going during longer tasks connected with questions we were already working through.

My first thought was how much of what I’ve been working on will we no longer need?

This is another great example of how harnesses are compressing into models. That's why always build your harness and repo that you can destroy every 6 months and build it from scratch again. It should be very fluid and modular, so you can easily bring the frontier capabilities. That's why building a duplex harness that happens to use ASR, STT, LLM, TTS models is not a good solution since duplex is commoditized today. Build a harness where the duplex model itself is a replaceable component. This is valid for any software field.

After spending months with these problems, I appreciate the work behind GPT-Live more deeply. Through client delegation, anyone can connect GPT-Live to existing harnesses, where the models you choose work with your tools.

Thank you to the OpenAI team. This is a big win for anyone building solutions to benefit humanity.

[OpenAI Developers @OpenAIDevs]

GPT-Live-1 is now available in the API.

Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose.

[Muratcan Koylan @muratcan]

@thaiscbranco_ Congrats 🎆🎇

[Muratcan Koylan @muratcan]

@bubbleboi Anyone who can’t split the G shouldn’t be making frontier policy.

[Muratcan Koylan @muratcan]

@DavidSacks Based

[Muratcan Koylan @muratcan]

@nickmmark This is very plausible

[Sara Hooker @sarahookr]

On the slow death of scaling.

[Muratcan Koylan @muratcan]

@aidangomez @jenningsgreg Thanks for not bowing down to the ai/industrial complex

[Muratcan Koylan @muratcan]

@cdngdev heard 🫡

[Muratcan Koylan @muratcan]

@JeremyKrak Our primary focus is currently on enterprise customers, but I’ll share this with the team, I'm sure they are happy to help.

[Muratcan Koylan @muratcan]

@dctanner @usetoyo I'm actually deeply happy that these technical engineering problems are being solved in voice ai, so now we can focus on bringing these capabilities in different mediums to more people, and create more solutions.