Amy Heineike

Bring on the algorithms! Founding AI Engineer at Tessl Previously @ PrimerAI, 7bridges.

[Amy Heineike @aheineike]

Very excited to be back in San Francisco for the week. We have a booth at @aiDotEngineer to talk about @tessl_io's tools for working with skills and using them in agentic dev systems, and also looking forward to seeing old friends.

[Amy Heineike @aheineike]

@ethanyish @AmplifyPartners @kleinerperkins @Theoryvc Congratulations Ethan!

[Amy Heineike @aheineike]

@pamelafox We're building a package manager for skills at https://t.co/rDboFmpRWd. Plus, you get evals built in, so you can check the effectiveness of the skill and how well structured it is.

[Guy Podjarny @guypod]

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!

[Amy Heineike @aheineike]

Here's the full blog post, with some links to how to install Tessl, find skills, run reviews and publish your own easily. Lots more coming that we'll share soon https://t.co/m76IOHq8ck

[Amy Heineike @aheineike]

Excited to share our latest release -- a proper package manager for agent skills with evals built in.

[Guy Podjarny @guypod]

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: https://t.co/GlGcc4QMRa

[Amy Heineike @aheineike]

Great work from my colleague @migorinova -- understanding how we help coding agents follow the actual APIs of libraries they depend on.

[Maria Gorinova @migorinova]

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!

https://t.co/3r7PX6A7Qy

[Amy Heineike @aheineike]

Excited for @aiDotEngineer Paris today. Really enjoyed the conversations at our booth so far -- working with coding agents, specs and the joy of evals.

[Simon Maple @sjmaple]

We've been busy here at @tessl_io, working on Spec-Driven Development. Some BIG news: We're announcing the release of two new products from Tessl https://t.co/k46SF4KoAp

The 𝐓𝐞𝐬𝐬𝐥 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤: makes agents capture intent in specs before coding, aligning you and the agent on what to build. This truly keeps agents on the rails - It’s available in private beta - ask for access on the site: https://t.co/TmIfR7CtsK

The 𝐓𝐞𝐬𝐬𝐥 𝐒𝐩𝐞𝐜 𝐑𝐞𝐠𝐢𝐬𝐭𝐫𝐲: helps agents use open source better. A registry of over 10,000 usage specs that guide agents in application development. It’s in open beta and free to use - check it out: https://t.co/HzTezL6TaU

It's been such a fun journey already, just to get to beta, with the best team I've ever worked with. Super excited sharing our products with the community so we can build in the open!

Want to see it in action? Here's @bgalbs showing off the products!

[Guy Podjarny @guypod]

I’m excited to launch @tessl_io's first products!
Introducing the Tessl Framework and Tessl Spec Registry, which integrate into any agent to keep it on rails and well informed using Spec-Driven Development.
More details in the launch post: https://t.co/qO5q0a6zhO

It’s a big milestone in the journey towards AI Native Development🚩
I’m proud of our amazing team, and keen to get beta feedback from the community! ❤️

What’s the problem we’re solving?
Agents are powerful, but they’re very unreliable. They hallucinate, claim false success and break things often enough that it’s hard - and tiring - to use them on production code.

How are we helping?
The Tessl Framework makes agents capture intent in specs before coding, aligning you and the agent on what to build. It adds tests as harder guardrails, and stores specs as long term memory of what your product should do. It’s available in private beta - visit our home page to request early access: https://t.co/sAeHKR5mv3

The Tessl Spec Registry helps agents use open source better. It contains over 10,000 usage specs for using libraries, which you can add to your project like regular dependencies. It also lets you distribute your own guidance and policies to agents. It’s in open beta and free to use! Check it out here: https://t.co/BC6UOcK3nl

Both products are just the beginning, and we’re committed to building them in the open. Check out the video or our launch blog post for more info: https://t.co/qO5q0a6zhO

Looking forward to hearing your thoughts, in the thread or our community Discord!

[Ben Galbraith @bgalbs]

🧵 AI coding agents — like Claude Code and Codex — are amazing, but they’re also unpredictable and hard to control.

Today at Tessl we’re announcing a framework that helps keep agents on the rails and arms with the context they need to get things right.

It has two parts:

[Amy Heineike @aheineike]

Its very early days, so we're excited to see where this goes, and improve it from here.

[Amy Heineike @aheineike]

Very excited that you can now try out what we've been building at Tessl:
A massive registry of usage specs, to steer coding agents to make better use of dependencies - now in public release.
An expansion of our closed beta for the Tessl framework for Spec Driven Development.

[Tessl @tessl_io]

Today marks a big milestone for Tessl!

The Spec Registry is now live in public beta with 10,000+ usage specs for open-source libraries.

Agents get the version-accurate APIs, examples, and history they need to use OSS reliably. No more hallucinations or endless fix loops. 🧵

[Amy Heineike @aheineike]

We had early access to Claude 4 to run through our benchmarks - and the results were so impressive - a huge jump up in performance. Congratulations to the team!

[Anthropic @AnthropicAI]

Introducing the next generation: Claude Opus 4 and Claude Sonnet 4.

Claude Opus 4 is our most powerful model yet, and the world’s best coding model.

Claude Sonnet 4 is a significant upgrade from its predecessor, delivering superior coding and reasoning.

[AI Native Dev @ainativedev]

@aheineike & @tessl_io's AI Engineering team ran a range of coding challenges comparing o3-mini & GPT-4.5.
Tl;dr:
o3-mini outperformed GPT-4.5, but no model rules AI-native dev—combining models’ strengths is the future.
Written by @FernandezBap: https://t.co/oqsnxcMpQ1

[Guy Podjarny @guypod]

I’m humbled by the vote of confidence in our team and mission from these amazing investors. It’s also true recognition of the amazing team we assembled to take this on.

Their work isn’t visible yet, but I’m regularly amazed by the depth and speed of thinking and execution in our founding team.

They’re a brilliant bunch, while also being fun, caring and driven. I’m massively grateful they’ve chosen to join this journey.

[Amy Heineike @aheineike]

Big funding news today. Fuel for the fire, and lots to build...

[Guy Podjarny @guypod]

Excited to share @tessl_io has raised $125M, to build out AI Native Software Development!

This includes a $25M Seed from April, led by @Boldstartvc & @GVteam, and a $100M Series A, led by @IndexVentures with @Accel, @GVteam and @Boldstartvc participating.

More info in our blog post announcement - https://t.co/X0OVmm7aPY

[Simon Maple @sjmaple]

Happy to announce AI Native Dev Con, a virtual conference about AI-powered development today and tomorrow, taking place on November 21st. The CFP is now open, and you can also register to save your spot

https://t.co/i5cULF4db1

[Simon Maple @sjmaple]

"If you took a software engineer from 1998 and dropped them into 2018, they wouldn't even understand what was happening. We're about to go through the same thing with AI-enabled software development." - Peter Guagenti #AINative #SoftwareDevelopment #AIDevelopment

[Amy Heineike @aheineike]

Very excited to share news about my new gig. Exploring the future of software development in the AI era.

[Guy Podjarny @guypod]

I’m excited to introduce my new company, @tessl_io!

We’re reimagining software development for the AI era, and helping shape a new software development paradigm we call AI Native Software Development.

The AI tools we use for development today are novel and powerful, and we should all invest in embracing them - but they are only the beginning. They optimize dev workflows created before the LLM revolution, bolting on AI at various spots.
To tap the true potential of AI-powered dev, we need to rethink these workflows from first principles, assuming AI is built in. This is what we mean by AI Native development.

(thread for more)