205 Photos and videos
Pinned Tweet
Jan 29
I'm excited to announce Tessl's 𝐩𝐚𝐜𝐤𝐚𝐠𝐞 𝐦𝐚𝐧𝐚𝐠𝐞𝐫 𝐟𝐨𝐫 𝐚𝐠𝐞𝐧𝐭 𝐬𝐤𝐢𝐥𝐥𝐬!
Install any skill, browse over 2,000 evaluated skills on the Tessl Registry, and be setup to manage those skills over time!
Find out more on our blog (link in the comments), or get started with:
npm i -g @ tessl/cli && tessl skill search
Agent skills are powerful, and need professional dev tools to match.
Don't copy skills to your repo - install them as a dependency you can update.
Don't blindly use a skill and hope it works - evaluate its quality and efficacy.
Don't hope someone notices you updated your skill - version it properly.
Tessl enables all of this and more for skills, as well as other forms of context.
Super excited to be launching this, and for the many more enhancements in the queue!
Explore Skills in Tessl Registry: tessl.co/dzm
1
5
18
4,131
Guy Podjarny retweeted
Jun 1
Do cloud canaries still work when the attacker isn't human?
We pointed 10 frontier models at AWS across 951 runs. 95.9% tripped a canary before any critical action - and telling models to expect deception reduced the number of accounts fully compromised from 20% to 3%.
Read the research here: agentic.tracebit.com
11
41
1,022,419
Grep just achieved SOTA on the three major deep research benchmarks, beating Perplexity, Google, Nvidia, OpenAI, and Anthropic.
We're a two-person founding team.
15
22
286
64,362
Guy Podjarny retweeted
Mar 25
Writing a SKILL.md is without testing it, is writing it blind. Ultimately, you don't know if the agent follows it, if parts of the skill are redundant, or if it even makes things worse.
We wrote a skill called skill-optimizer which solves this problem through running structured evals, comparing agent performance with and without the skill, and giving a clear score delta. It combines two approaches: a static review of the skill instructions, and real task-based evaluations that simulate realistic scenarios and grade outcomes.
I used a real Fastify skill example, which @matteocollina wrote and identified regressions, diagnoses issues, applies fixes, and verifies improvements automatically, turning a 67% success rate into 94%.
PR on it's way, Matteo! tessl.io/blog/stop-guessing-…
3
9
1,305
Guy Podjarny retweeted
Mar 26
.@chadfowler replaced 70% of his codebase in 3 months and cut costs by 75%.
His #1 rule: don't write a service longer than a page.
The shorter it is, the easier it is to replace.
Mar 24
"The code that we have is a liability. The system is the asset we're building."
@chadfowler, VC at Blue Yard Capital (@blueyard) and former CTO at Wunderlist, sits down with @guypod to discuss the Phoenix Architecture: software designed to be replaced rather than maintained.
In this episode:
• why was the code written by Chad never longer than a page
• how he replaced 70% of a codebase in 3 months and cut costs by 75%
• shipping AI code no human ever reviewed, and how to make it safe
• the shadow specs your agents are making without you
• why your system should work with the worst LLM, not just the best
If you're still thinking about your codebase the old way, this one will change that.
(0:00) Trailer
(1:07) AI DevCon
(2:01) Introduction
(3:41) Origin story: euthanising legacy systems
(5:45) Immutable infrastructure as inspiration
(6:48) Disposable software and immutable code
(9:00) Cattle versus pets for code
(10:03) Making disposable code feasible at Wunderlist
(12:31) Phoenix Architecture
(15:16) Extreme programming lesson: do hard things constantly
(17:04) What level of detail should specs have?
(19:15) Pace layers and stable regeneration
(22:37) New programming languages versus patterns
(29:47) Compiling to system architectures
(30:45) Training the programmer versus defining the system
(35:03) Personalised and malleable software
(37:48) Local first and shared data models
(45:08) Evaluations as the real codebase
(49:36) Testing the agent versus testing the system
(55:38) Path of adoption
(01:00:48) Wrap-up
1
4
20
4,694
Guy Podjarny retweeted
Mar 19
My main takeaway is that skills are software, and the same rules apply. The things that make a bad skill also make a bad software component, eg. being badly scoped. "Should we factor this out" has become "Should we make a skill for this". Same stuff, different form factor
2
5
3
1,068
Guy Podjarny retweeted
Feb 26
Snyk founder is working on something new 👀
Feb 26
Agent skills help agents use your products, build in your codebase and enforce your policies. They're the new unit of software for devs - but most are still treated like simple markdown files copied between repos with no versioning, no quality signal, no updates.
Without AI evaluations, you can’t tell if a skill helps, provides minimal uplift or even degrades functionality. You spend your time course-correcting agents instead of shipping.
@tessl_io is a development platform and package manager for agent skills. Today, I’m excited to launch on Product Hunt and announce that you can evaluate your skill and optimize them on Tessl. This means you can stop debugging agent output and start shipping quality code, faster. Real example: we've helped ElevenLabs ship skills that double agent success in using their APIs.
If you're building a personal project, maintaining an OSS library, or developing with AI at work, you can now evaluate your skill and optimize it to help agents use it properly. Check us out on Product Hunt. If it’s useful, we’d love your upvote - and even more, your feedback in the comments: producthunt.com/products/tes…
1
1
4
747
Feb 26
Agent skills help agents use your products, build in your codebase and enforce your policies. They're the new unit of software for devs - but most are still treated like simple markdown files copied between repos with no versioning, no quality signal, no updates.
Without AI evaluations, you can’t tell if a skill helps, provides minimal uplift or even degrades functionality. You spend your time course-correcting agents instead of shipping.
@tessl_io is a development platform and package manager for agent skills. Today, I’m excited to launch on Product Hunt and announce that you can evaluate your skill and optimize them on Tessl. This means you can stop debugging agent output and start shipping quality code, faster. Real example: we've helped ElevenLabs ship skills that double agent success in using their APIs.
If you're building a personal project, maintaining an OSS library, or developing with AI at work, you can now evaluate your skill and optimize it to help agents use it properly. Check us out on Product Hunt. If it’s useful, we’d love your upvote - and even more, your feedback in the comments: producthunt.com/products/tes…
1
10
1,283
If agentic development is the future, then skills are the atomic unit.
But how do you move from experimental to production-grade agents?
You need @tessl_io, the dev-grade package manager for skills. It’s your registry for evaluated skills platform to manage their full lifecycle.
Congrats @guypod & the team! 🚀
Feb 17
Agent skills help agents use your products, build in your codebase and enforce your policies. They’re not just words - they are what the unit of software for agentic devs, and need powerful dev tools to match. That is what @tessl_io offers.
Tessl is the package manager and development platform for skills. It offers a full dev lifecycle, helping you generate, evaluate, distribute and observe skills & context, developing them to the professional grade they warrant.
Today, I’m excited to announce the general availability of our task evals, which help you understand how good your skills are. Such insight is critical to making your skills great, avoiding regression, and applying learnings from their real world usage.
For example: @Cisco's software-security skill shows a 1.8X improvement in securing coding in its benchmark, and @ElevenLabs's agents skill boosts success by almost 3X!
However, not to name names, we often see skills that provide minimal uplift while consuming context window space, or even degrade functionality.
As Spencer Kimball, CEO of Cockroach Labs, put it when we shared early versions of this: evaluation is what makes agentic coding outcomes converge instead of drifting.
Task evals are joining a long list of powerful context development tools, such as:
* Review skills against quality best practices
* Generate and maintain skills and docs for using your libraries & platform
* Distribute versioned skills to your dev team and ecosystem
* Consume skills easily and safely, and keep them up-to-date
Skills are a central part of software development. If you’re serious about making agentic dev successful in your org, or helping your customers’s agents use your products, you need to invest in them. We hope Tessl can help.
Check out links in the thread to get started!
3
2
6
1,692
Guy Podjarny retweeted
Feb 17
new product to help agents gain new skills, and evaluate their skills...from the @guypod and @tessl_io team...
Feb 17
Agent skills help agents use your products, build in your codebase and enforce your policies. They’re not just words - they are what the unit of software for agentic devs, and need powerful dev tools to match. That is what @tessl_io offers.
Tessl is the package manager and development platform for skills. It offers a full dev lifecycle, helping you generate, evaluate, distribute and observe skills & context, developing them to the professional grade they warrant.
Today, I’m excited to announce the general availability of our task evals, which help you understand how good your skills are. Such insight is critical to making your skills great, avoiding regression, and applying learnings from their real world usage.
For example: @Cisco's software-security skill shows a 1.8X improvement in securing coding in its benchmark, and @ElevenLabs's agents skill boosts success by almost 3X!
However, not to name names, we often see skills that provide minimal uplift while consuming context window space, or even degrade functionality.
As Spencer Kimball, CEO of Cockroach Labs, put it when we shared early versions of this: evaluation is what makes agentic coding outcomes converge instead of drifting.
Task evals are joining a long list of powerful context development tools, such as:
* Review skills against quality best practices
* Generate and maintain skills and docs for using your libraries & platform
* Distribute versioned skills to your dev team and ecosystem
* Consume skills easily and safely, and keep them up-to-date
Skills are a central part of software development. If you’re serious about making agentic dev successful in your org, or helping your customers’s agents use your products, you need to invest in them. We hope Tessl can help.
Check out links in the thread to get started!
1
1
7
6,012
Guy Podjarny retweeted
Feb 17
Jobs called computers "bicycles for the mind" -- tools we could shape to our will.
But they never were. Until now.
Every morning an agent preps me for my day -- calendar, news, last 24hrs of Slack -- in a personal podcast. I made it by asking. Same for hundreds of other things.
Launching @dreamer in beta today.
That 🧠 bicycle, finally.
dreamer.com
Introducing Dreamer.
A place to discover, build, and enjoy agentic apps.
It’s your home for personal intelligence.
Now in beta. Sign up👇
86
71
448
229,491
Feb 17
Agent skills help agents use your products, build in your codebase and enforce your policies. They’re not just words - they are what the unit of software for agentic devs, and need powerful dev tools to match. That is what @tessl_io offers.
Tessl is the package manager and development platform for skills. It offers a full dev lifecycle, helping you generate, evaluate, distribute and observe skills & context, developing them to the professional grade they warrant.
Today, I’m excited to announce the general availability of our task evals, which help you understand how good your skills are. Such insight is critical to making your skills great, avoiding regression, and applying learnings from their real world usage.
For example: @Cisco's software-security skill shows a 1.8X improvement in securing coding in its benchmark, and @ElevenLabs's agents skill boosts success by almost 3X!
However, not to name names, we often see skills that provide minimal uplift while consuming context window space, or even degrade functionality.
As Spencer Kimball, CEO of Cockroach Labs, put it when we shared early versions of this: evaluation is what makes agentic coding outcomes converge instead of drifting.
Task evals are joining a long list of powerful context development tools, such as:
* Review skills against quality best practices
* Generate and maintain skills and docs for using your libraries & platform
* Distribute versioned skills to your dev team and ecosystem
* Consume skills easily and safely, and keep them up-to-date
Skills are a central part of software development. If you’re serious about making agentic dev successful in your org, or helping your customers’s agents use your products, you need to invest in them. We hope Tessl can help.
Check out links in the thread to get started!
3
9
29
20,117
Feb 17
Find skills in the Tessl Registry: tessl.io/registry
Evalute your skills - see our docs: docs.tessl.io/evaluate/evalu…
Browse the mentioned skills:
- Cisco: tessl.io/registry/cisco/soft…
- ElevenLabs: tessl.io/registry/skills/git…
256
Feb 11
What does an Observability solution built for agents look like? Do they need dashboards? Do they care about log formats? And how does such a product interact with humans?
I had a fascinating conversation about that and more with @mirko_novakovic on the @ainativedev . We discussed how :
- LLMs natively understand OpenTelemetry
- Humans like dashboards but agents like text
- Context is key to making agents work
- Use cases that work today with agents
Throughout, we had an open conversation challenging what is an observability product if you standardize the format, rely on 3rd party LLM analysis and take away the dashboard. Spoiler alert - there's much value to deliver, but it’s different!
Mirko founded @dash0hq, an AI Native O11y company, and previously built an 011y company called Instana, and sold it to IBM. He knows the space :)
Great conversation with a superb guest - a must listen!
Full episode here: tessl.co/swe
1
6
557
Agent skills are getting harder to manage.
Most teams still treat skills as static artifacts: markdown files, created or copied from repo to repo.
This quickly leads to debt, with skills growing stale and copies falling out of sync.
That's where Tessl comes in!
Today, we launched agent skills on Tessl 🎉
Tessl lets you treat skills like software, not snippets:
- Discover evaluated skills in the Tessl Registry
- Install and evaluate skills via CLI or from GitHub
- See how skills perform across agents and models
- Version, update, and evolve skills safely over time
Now you can discover evaluated skills in the Tessl Registry or install and test any skill from GitHub.
💻 npm i -g @tessl/cli && tessl skill search
🔗 Explore Tessl Registry: tessl.co/cds
#agentskills #devtools #aidevtools
2
10
1,075
Guy Podjarny retweeted
Jan 26
Today marks the next chapter for @synthesiaIO as we announce our $200M Series E at $4B led by GV.
It’s been quite a ride since 2017 when we set out to transform how people make video. Now, 8 years later, so much of that vision has come to fruition as evident everywhere around us. It’s truly incredible how good AI video has become in the last two years.
Up until now it’s been mostly about making video as we know it with AI. This is often referred to as the ‘bridge-period’ where new technology copies old formats, like when the first films were just recorded theater or early GPS was just a digital map with no navigation.
Now, it’s time to figure out what AI-native video truly looks like - when we let go of the priors and rethink video in the context of language models, real-time video generation, smartphones and so much more.
With this new round, it’s all about helping people work better – both with AI video and a bunch of new real-time products, like Skills. We’ll be sharing much more in the coming months!
Grateful to all of our amazing customers, Synthesian’s and investors!
11
12
64
15,741
Guy Podjarny retweeted
20 Nov 2025
Super excited to share what we've been doing at @tessl_io to improve the quality of code generated by AI agents!
We introduce a new way to measure abstraction adherence and show how Tessl 's usage specs significantly boost it
Check out the full article!
tessl.io/blog/proposed-evalu…
1
2
4
847
Guy Podjarny retweeted
12 Nov 2025
Cloud costs are eating software margins alive. Today we're announcing Frugal's 5M seed round. Frugal is the first Application Cost Engineering (ACE) platform: fixing inefficient code before it hits production. frugal.co #finops #ai #CloudCostOptimization
1
3
635
Guy Podjarny retweeted
14 Oct 2025
Introducing Capture to Notion.
One ⌘ shortcut to capture anything straight to Notion.
Your new w̶e̶b̶ universal clipper.
26
41
370
53,308
Guy Podjarny retweeted
7 Oct 2025
What if AI could extend your app’s capabilities?
Steve Manuel (@nilslice), founder and CEO of @dylibso, joins @sjmaple on AI Native Dev to decode MCP and reveal how mcp.run lets developers safely scale AI beyond its original limits.
He dives into how developers can extend AI beyond text, connect to diverse data sources, and design enterprise-grade workflows with confidence, all with MCP.
Catch their full conversation now. Link in the comments.
2
6
5
1,208