Scope (YC P26)

Make your product discoverable and usable by every AI agent.

[Scope (YC P26) @tryscope_app]

In our latest tests, agents built their own solutions instead of using an existing product nearly 35% of the time.

And even when explicitly told which product to use, 28% of runs still failed.

Agent experience is becoming a new problem for software companies. Here's where things break ↓

[Lin @lindictive]

AI agents skip vendors one time in three and fail 28% of the time even when told which product to use. Here's where your product loses them, with data. https://t.co/jRHIx1QYq0

[Anand Paj @anandPa94]

How good are AI Agents at using your tool?
Find out with Scope: https://t.co/s5nyyphvoL

We partnered with @AgentMail to test 25 model and harness combinations across 7 email tasks.

AgentMail co-founder @adisingh shares why they needed an eval in the first place and why they chose Scope to run it 👇

[Scope (YC P26) @tryscope_app]

Journeys is now live! Track AI Agent Attribution now.

https://t.co/BpGOAszGC8

[Anand Paj @anandPa94]

We just solved attribution for AI Agents.
Introducing Journeys.

AI agents are visiting your product. Now you can see what they’re trying to do, what converts, and what’s making them churn.

With Journeys, you can:
- See the moment AI agents visit your product
- Follow the paths they take
- Track purchases, installs, and completed setup
- Pinpoint where agents get stuck and whats blocking them
- Review fixes prepared for you and publish them with a click
- Watch how those changes improve conversion

Turn agent traffic into revenue.
Try Journeys at https://t.co/S6pve0LkjZ

[Anand Paj @anandPa94]

We just solved attribution for AI Agents.
Introducing Journeys.

AI agents are visiting your product. Now you can see what they’re trying to do, what converts, and what’s making them churn.

With Journeys, you can:
- See the moment AI agents visit your product
- Follow the paths they take
- Track purchases, installs, and completed setup
- Pinpoint where agents get stuck and whats blocking them
- Review fixes prepared for you and publish them with a click
- Watch how those changes improve conversion

Turn agent traffic into revenue.
Try Journeys at https://t.co/S6pve0LkjZ

[Adi Singh @adisingh]

Introducing AgentMail Evals.

We tested how AI agents use tools: 25 model and harness combinations, seven email tasks, with AgentMail as the proxy.

[Lin @lindictive]

excited to have worked on this with the @tryscope_app and @agentmail team!

the same model can perform very differently depending on the harness & tools it uses. product-specific evals give builders a much more detailed image of what is working and what isn't

[Lin @lindictive]

Your customers aren't human anymore, they're AI agents. Here's why your product will never succeed unless it's built for them. https://t.co/XTUq19B5Bv

[Lin @lindictive]

You can now A/B test how AI agents discover and use your company, instantly, before anything hits production.

In this first episode, my cat and I will show you how to test an agent's reaction to any piece of content you’re considering publishing, before you actually ship it.

See what changes their behavior, what they understand, and what makes them more likely to succeed.

Want to try it out? Comment below or check out https://t.co/OE2iGF1xLf

[Lin @lindictive]

You lose customers every time an agent like Codex or Claude Code uses your product and fails to complete a task.

At @tryscope_app we built a feature to find those errors for you and fix them. Let’s use @vercel as an example to show you how it works!

Want us to run a workflow for you? Drop it in the comments below :)

[Lin @lindictive]

the best competitive gamers make the best builders. same grind & obsession, refusal to give up, and attention to detail. i'd bet on those people every time.

sometime after ai blew up, a lot of my friends and I traded gaming for tech. but gaming was how we stayed close across time zones & allowed us to catch up and hangout until 5am.

this weekend @nicole_clash and I are bringing it back for one day. the sharpest founders, engineers, and operators in SF, in one room, playing the games that we love.

come make friends, meet really smart ppl, and maybe leave with some cash.

apply below:
https://t.co/yWRcKjAl3N

[Scope (YC P26) @tryscope_app]

@thru_xyz @lindictive @PlanetScale @Medal_TV @joinef

[jen @jenfang]

details @tryscope_app & @sfcompute ikebana event 💐

@lindictive @emolyonx

[jen @jenfang]

Scope @tryscope_app postcard en route

using @figma motion & @buburdin https://t.co/T6meeux99e

[Anand Paj @anandPa94]

Four weeks ago, we pivoted. Yesterday was Demo Day.

I was honestly nervous about how to tell investors. The product was moving fast. Some parts were rough. Some features were still in beta. Most things were not perfect. But being upfront ended up being one of the best things we did. When something is not perfect, just say it.“This feature is in beta. It is not perfect yet. But it is getting better every day, and in a few weeks it is going to be killer.” That kind of honesty works when it is backed by speed.

We talk to customers nonstop. We ship like crazy. We learn fast. Every week, the product gets better because the market is telling us exactly what to build. You do not need to be perfect to be impressive. You need to know what is true. Own what you know. Own what you do not know. Then show how fast the gap is closing.

Since the pivot, we went from 0 to $30k MRR. Last week, we became profitable.
Don’t be scared to pivot.

[Anand Paj @anandPa94]

I’m part of the YC P26 batch that got offered $2M by Sam Altman in funding with OpenAI tokens. I’ve had a few days to think about it so I thought I would share how I’m thinking through the entire thing.

The first thing that makes this offer interesting is how founder-friendly the deal is, meaning that it is low dilution and is a huge token allocation for a small team.
If you are spending thousands of dollars of token per month building, this could be a huge relief and can decrease your burn and improve your startup life expectancy.

There are a few questions that I am asking myself regarding this deal: do I want to lock in with one provider? If we start building an entire company's workflow on OpenAI, what will be the switching cost in a few years? How is token cost gonna evolve and how will that change the value of the deal? Do I need cash right now?

On the other hand, if you are building an AI-native company this could be the opportunity to keep pushing. It is part of the scaling law, the more you put resources in it, the more the potential (at least it kinda worked for compute and LLMs). It also gives you leverage while fundraising because now the question becomes “why would I raise at a X valuation when I can have all this spend already on tokens?” By having that alternative option you can really think about which investors are right for you, and what they can bring to the table aside from just capital.

In the end it obviously depends on your situation and there is no playbook here. I’m excited to see how this will turn out to be in 3, 5 or 10 years from now. This is the beginning of tokens becoming a utility.

[Scope (YC P26) @tryscope_app]

@anandPa94 👋 come say hi

[Anand Paj @anandPa94]

Book a demo here: https://t.co/EvDxgGeu7V

[Anand Paj @anandPa94]

Introducing Scope (@tryscope_app): the platform that breaks down exactly how AI agents see and interact with your product.

Products are increasingly being used through Claude Code, Codex, Cursor, and similar AI agents. Most companies have no idea what an agent's decisions look like. Scope fixes that.

What Scope does:

→ Run real workflows across Claude Code, Codex, and Cursor, and show you when agents choose you vs. a competitor
→ Pinpoint exactly where agents get stuck: auth, onboarding, docs, setup
→ Show the full trace of every agent decision so you know what broke and why
→ Tell you what to change to get picked more, used more, and kept more.

Every company we've shown this to has immediately thought of a workflow they want to run.

Want to know how agents see your product? ↓

[Y Combinator @ycombinator]

Scope (@tryscope_app) helps companies see when AI agents choose them, get stuck, or pick a competitor, and what to change to improve it. It runs real workflows across Claude Code, Codex, Cursor, and similar AI agents.

Congrats on the launch, @anandPa94!

https://t.co/jcapB01zk5

[Anand Paj @anandPa94]

I had a working product, ~10 paying customers, had just signed a Fortune 10 logo, and last week I still decided to pivot. 1 week later, I know it was the right call.

Quick retro. On Wednesday 10 days ago, I was in a demo with a prospect who just raised a Series B. It was going well until I asked about next steps. The response were "let me check with marketing if we have budget for this" and "we're a small team right now." But this wasn't the first time I was hearing that. The pattern was clear, my product was great but not urgent. A box to check, maybe market timing, who knows.

So mid-call, I started asking about a slightly adjacent idea that a few earlier prospects had hinted at. They were on classic GEO/AEO (humans discovering products through AI), but kept mentioning how AI agents themselves were already discovering and choosing for them. Different problem, and maybe a bigger one.

And then I felt it shifting. The same prospect who 10 minutes before was saying "idk, small team, no budget" went to "let me send this to XX right now, they'll take it from here." I met with that person the next day.

It's never easy to pivot from a working product with paying customers. But that shift in 30 minutes made me realize something bigger and more urgent might be there. So I gave myself 48 hours to talk to as many companies as I could that might be feeling this same pain, and go sell to them. My take: if someone went from "idk" to "we need this" in a 30 minutes call, I should be able to find at least 10 companies in 48 hours saying the same.

I cooked up a quick MVP and met with 13 founders and engineers I knew would potentially feel this pain. I closed 2 of them with one being an enterprise. The pull was something else compared to my previous product. Prospects were moving 1000x faster than anyone before so I knew I had to do it.

The vision was the only clear thing. The "how" was completely blurry though but I kept sketching and started building a v1.

1 week later, I have 12k MRR secured and the product is already in demo and showing value.

Now I'm facing a new problem: how to balance market pull and servicing my new customers extremely well. For now I'm thinking about focusing on a small set of paying customers and to service them extremely well. This will help me figure out the wow moment and the repeatable cycle, then open it up and get aggressive on growth. I will tell you more next week after experimenting :))