Is Scope profitable or generating revenue in 2026?

No public revenue has been disclosed. Scope launched via Y Combinator on 2026-05-11 with a team of 2 and named only one confirmed customer, Blaxel, plus one unnamed 'decacorn' client. No revenue, growth rate, or pricing figures have been published as of July 2026.

Could AI labs like Anthropic or OpenAI replace what Scope does?

Yes, plausibly. Scope's core mechanic, running agents against a product and logging failure traces, is exactly what Anthropic or OpenAI could ship natively into Claude Code or Codex at zero marginal cost, since Anthropic already owns the underlying MCP protocol introduced in late 2024.

Who are Scope's biggest competitors?

The Prompting Company, a same-batch YC startup with near-identical positioning ('help products get discovered & used by AI agents'), is Scope's closest direct rival as of May 2026. Scrunch AI and the better-funded Profound also compete in the adjacent agent-experience and AEO space.

Scope is a private, founder-owned startup backed by Y Combinator, part of its Spring/2026 batch (YC P26), with Nicolas Dessaigne listed as its YC Primary Partner as of the May 2026 launch. No external lead investor or valuation has been disclosed.

What does Scope's product actually do?

Scope runs live AI coding agents (Claude Code, Codex, Cursor-style) against a customer's product, capturing traces of tool calls, errors, and reasoning to identify friction points like unclear docs or broken auth flows, then delivers a specific fix list, per its tryscope.app site as of May 2026.

How many customers does Scope have?

As of its May 2026 YC launch, Scope names one confirmed customer, Blaxel (an AI agent hosting infrastructure startup), and references one unnamed customer described as a 'decacorn.' No customer count or growth figures beyond these two have been published.

YC Spring 2026 · United States

Scope

Scope tests how AI agents use a company's software and flags where agents fail.

Losing its protection

Assessed 27 July 2026 · Evidence as of 27 July 2026

The structural protection is eroding.

On the recordSince 27 July 2026 · Checkable from by late 2027 · Every call on the record Checkpoint: Estimated window for Anthropic or OpenAI to ship native agent-readiness diagnostics into Claude Code or Codex

The call

Scope is two people chasing a real problem in a category that frontier labs and analytics incumbents could both absorb before it compounds any data or lock-in.

Scope is a young company that tests AI agents to find what breaks and why. Right now it's hard to tell if this is a real business or just a useful tool, because Anthropic (the company behind Claude) and OpenAI both have the power to add these same testing features directly into their AI platforms for free. Scope has almost no customers yet and one rival doing the exact same thing at the same stage. The only thing that could save Scope is if it builds up a huge library of data about how agents fail across thousands of different products and locks customers into using it as part of their regular software development process, but neither of those things exist yet.

Why

AI pressure: Anthropic and OpenAI (native agent diagnostics inside Claude Code, Codex). Frontier labs already run agents against code; adding failure-diagnostics natively could fold Scope's function into the platform for free. Could AI do the core job? Yes, likely.

Bigger players: Ahrefs, Semrush, Profound (SEO/analytics platforms adding agent-visibility features). Established analytics vendors with existing customers and distribution are adding agent-readiness checks as a feature, not a company. Could a giant copy it? Yes, easily.

What protects it. None built yet: the founder's interpretability and GEO background helps read agent failures, but no dataset or lock-in exists today.

What would change this call

Anthropic or OpenAI ships native agent-diagnostics into Claude Code or Codex within 18 months. Would show the core feature is being absorbed by the platform layer, undercutting Scope's reason to exist.
Scope names a second and third paying customer beyond Blaxel within 90 days. Would show repeatable buyer motion rather than a one-off pilot.
The Prompting Company announces funding or named customer wins before Scope does. Would show a same-stage rival winning the identical pitch and buyer first.

Key numbers

Team size: 2 employees (2026-05-11)
2 customers: Blaxel plus 1 unnamed decacorn (2026-05-11)
YC batch: Spring 2026 launch (2026-05-11)

Named risks

Anthropic or OpenAI could natively ship agent-diagnostic telemetry into Claude Code or Codex, obsoleting Scope's core feature. Severity: high
Team of only 2 people with no disclosed funding beyond standard YC terms limits execution capacity. Severity: high
Same-stage, near-identical competitor The Prompting Company could out-execute on identical thesis. Severity: medium
No proprietary data moat or CI/CD workflow lock-in built yet, leaving product easily replicable by adjacents like Ahrefs or Semrush. Severity: medium
Only two named customers as of May 2026 launch, unverified independently, indicating minimal proven traction. Severity: medium

Rivals, as read in July 2026

Rival	Direction	Evidence
The Prompting Company	contesting	Same YC batch, near-identical pitch, direct head-to-head competitor as of 2026-05-11
Scrunch AI	contesting	Also markets itself as 'Agent Experience Platform', positioned enterprise-grade as of 2026-06-15
Profound	losing to	Better-funded (vendor-claimed ~$96M Series C) with existing Agents/MCP feature set, per 2026-06-23 comparison
Ahrefs	losing to	Legacy SEO giant adding AI-visibility features (Brand Radar) that could extend into agent testing
Semrush	losing to	Bolting AI Visibility Toolkit onto existing large customer base as of 2026-06-22
Anthropic	losing to	Owns MCP protocol substrate; could ship native agent-diagnostics into Claude Code, threatening Scope's core feature

Questions people ask

Is Scope profitable or generating revenue in 2026?: No public revenue has been disclosed. Scope launched via Y Combinator on 2026-05-11 with a team of 2 and named only one confirmed customer, Blaxel, plus one unnamed 'decacorn' client. No revenue, growth rate, or pricing figures have been published as of July 2026.
Could AI labs like Anthropic or OpenAI replace what Scope does?: Yes, plausibly. Scope's core mechanic, running agents against a product and logging failure traces, is exactly what Anthropic or OpenAI could ship natively into Claude Code or Codex at zero marginal cost, since Anthropic already owns the underlying MCP protocol introduced in late 2024.
Who are Scope's biggest competitors?: The Prompting Company, a same-batch YC startup with near-identical positioning ('help products get discovered & used by AI agents'), is Scope's closest direct rival as of May 2026. Scrunch AI and the better-funded Profound also compete in the adjacent agent-experience and AEO space.
Who owns Scope?: Scope is a private, founder-owned startup backed by Y Combinator, part of its Spring/2026 batch (YC P26), with Nicolas Dessaigne listed as its YC Primary Partner as of the May 2026 launch. No external lead investor or valuation has been disclosed.
What does Scope's product actually do?: Scope runs live AI coding agents (Claude Code, Codex, Cursor-style) against a customer's product, capturing traces of tool calls, errors, and reasoning to identify friction points like unclear docs or broken auth flows, then delivers a specific fix list, per its tryscope.app site as of May 2026.
How many customers does Scope have?: As of its May 2026 YC launch, Scope names one confirmed customer, Blaxel (an AI agent hosting infrastructure startup), and references one unnamed customer described as a 'decacorn.' No customer count or growth figures beyond these two have been published.

Timeline

2026-06-23Third-party sites describe AEO/agent-experience category as splintering into sub-segments monthly
2026-06-15Scrunch AI marketed itself as an 'Agent Experience Platform' competing with Scope
2026-05-25Unrelated London-based company also named Scope raises $20M Series A (different entity)
2026-05-11Scope publicly launches via Y Combinator as 'The Agent Experience Platform'
2026-04-30YC Summer 2026 RFS names 'headless Software for Agents' as a theme, validating Scope's thesis
2025Scope founded in San Francisco
2024-lateAnthropic introduces Model Context Protocol, the substrate underlying Scope's product

Revision history

27 July 2026

First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.

In the same band

Trellis Surtr Defense Systems HeyClicky