Is Clawvisor profitable in 2026?

No revenue figures are disclosed. Clawvisor's homepage shows an illustrative usage example of '5 tasks, $36.00' per week, which is marketing copy, not verified revenue. As of July 2026 it has no named enterprise customers, making profitability unconfirmed and unlikely at this pre-revenue stage.

Could AI or Anthropic replace Clawvisor?

Yes, this is a real risk. Anthropic co-authored the MCP authorization spec alongside Microsoft and Okta/Auth0 as of May 13, 2026, and owns Claude Code, Clawvisor's flagship integration. If Anthropic ships native task-scoped runtime authorization, it could bundle away Clawvisor's standalone gateway.

Who is Clawvisor's biggest competitor?

Microsoft's Entra Agent ID and Descope's Agentic Identity Control Plane (shipped August 2025) are the most direct threats, since both are funded, shipped products in the same agent-authorization category. Arcade.dev, Permit.io, and WorkOS are also named competitors as of mid 2026.

Clawvisor is privately founder-owned by Eric Levine, who previously co-founded Berbix (YC S18, acquired by Socure in 2023 for a reported $70M) and led Airbnb's Trust and Safety engineering team. No external investors or parent company are disclosed as of July 2026.

Has Clawvisor raised a funding round?

No seed round, investor names, or valuation have been publicly disclosed as of July 21, 2026, per available sources. The company launched via Y Combinator's Spring 2026 (P26) batch on May 7, 2026, but Demo Day funding outcomes had not been reported.

What does Clawvisor actually do?

Clawvisor is a self-hostable, open-core Go-based security gateway that sits between AI agents (like Claude Code) and tools such as Gmail, Slack, and Stripe. It enforces task-scoped access, injects short-lived credentials, and logs an audit trail, per its GitHub repo dated May 16, 2026.

YC Spring 2026 · United States

Clawvisor

Gateway that authorizes and audits what AI agents can do, task by task.

Protection under test

Assessed 27 July 2026 · Evidence as of 27 July 2026

Something worth protecting exists and is being tested right now.

On the recordSince 27 July 2026 · Checkable from late 2026 · Every call on the record Checkpoint: YC S26 Demo Day and whether native runtime policy enforcement ships in MCP or Claude Code

The call

Clawvisor built a working gateway fast, but the same authorization layer is being drawn into Microsoft's and Anthropic's own products.

Clawvisor makes software that controls what tasks an AI agent is allowed to do, like a security guard for AI systems. The problem is that Microsoft, Okta, and other big identity companies are already building this same capability into their products, and Anthropic (the company behind Claude) is writing the technical standard that makes Clawvisor's approach copyable by anyone. Clawvisor could survive if it proves enterprise customers genuinely need a standalone product instead of using what comes built into Microsoft or Claude, but if the big companies ship native task controls into their AI platforms within six months, there may not be room for an independent company in this space.

Why

AI pressure: Anthropic (Claude Code native MCP authorization). Anthropic co-authored the MCP auth spec and could bundle task-scoped authorization straight into Claude Code, cutting out the gateway. Could AI do the core job? Yes, likely.

Bigger players: Microsoft Entra Agent ID, Okta, WorkOS. Microsoft, Okta, and WorkOS sell identity already and can add agent authorization as a feature, not a new sale. Could a giant copy it? Yes, easily.

What protects it. One founder with a Berbix exit and Airbnb trust and safety background, plus early support across five agent platforms.

What would change this call

Anthropic or Microsoft ship native task-scoped runtime policy enforcement into MCP or Claude Code within 6-12 months. Would show the standalone-gateway niche Clawvisor occupies is being absorbed by the platform.
Clawvisor discloses a funding round or names paying design partners in fintech, healthcare or defense by end of 2026. Would show real enterprise traction rather than a solo-founder pre-revenue project.
Clawvisor's YC team-size listing still shows 1 person by end of 2026. Would indicate stalled hiring and limited commercial momentum despite YC exposure.

Key numbers

Team size: 1, per YC company page (2026)
Founder's prior startup Berbix sold for ~$70M in 2023 (2023)
No disclosed funding round as of July 2026 (2026-07-21)

Named risks

No proprietary data, regulatory lock-in, or shipped differentiation beyond copyable architecture. Severity: high
Solo-founder team of one entering a crowded, well-funded category. Severity: high
Anthropic and Microsoft could bundle native task-scoped authorization into agent runtimes, closing the niche. Severity: high
No disclosed funding, revenue, or named enterprise customers as of July 2026. Severity: medium

Rivals, as read in July 2026

Rival	Direction	Evidence
Microsoft (Entra Agent ID)	gaining on	Shipped native agent identity/authorization product, noted June 6, 2026
Anthropic	gaining on	Co-authored MCP authorization spec with Microsoft and Okta/Auth0, May 13, 2026
Descope	gaining on	Shipped Agentic Identity Control Plane in August 2025, ahead of Clawvisor's May 2026 launch
Okta/Auth0	gaining on	Co-authored MCP auth spec; existing IAM business gives natural upsell path
Arcade.dev	contesting	Positions itself as agent-native action runtime, published rival category comparison July 10, 2026
Permit.io	contesting	Offers competing MCP Gateway and policy layer, active blog output through mid 2026

Questions people ask

Is Clawvisor profitable in 2026?: No revenue figures are disclosed. Clawvisor's homepage shows an illustrative usage example of '5 tasks, $36.00' per week, which is marketing copy, not verified revenue. As of July 2026 it has no named enterprise customers, making profitability unconfirmed and unlikely at this pre-revenue stage.
Could AI or Anthropic replace Clawvisor?: Yes, this is a real risk. Anthropic co-authored the MCP authorization spec alongside Microsoft and Okta/Auth0 as of May 13, 2026, and owns Claude Code, Clawvisor's flagship integration. If Anthropic ships native task-scoped runtime authorization, it could bundle away Clawvisor's standalone gateway.
Who is Clawvisor's biggest competitor?: Microsoft's Entra Agent ID and Descope's Agentic Identity Control Plane (shipped August 2025) are the most direct threats, since both are funded, shipped products in the same agent-authorization category. Arcade.dev, Permit.io, and WorkOS are also named competitors as of mid 2026.
Who owns Clawvisor?: Clawvisor is privately founder-owned by Eric Levine, who previously co-founded Berbix (YC S18, acquired by Socure in 2023 for a reported $70M) and led Airbnb's Trust and Safety engineering team. No external investors or parent company are disclosed as of July 2026.
Has Clawvisor raised a funding round?: No seed round, investor names, or valuation have been publicly disclosed as of July 21, 2026, per available sources. The company launched via Y Combinator's Spring 2026 (P26) batch on May 7, 2026, but Demo Day funding outcomes had not been reported.
What does Clawvisor actually do?: Clawvisor is a self-hostable, open-core Go-based security gateway that sits between AI agents (like Claude Code) and tools such as Gmail, Slack, and Stripe. It enforces task-scoped access, injects short-lived credentials, and logs an audit trail, per its GitHub repo dated May 16, 2026.

Timeline

2026-07-21Listed on APIs.io directory as AI agent authorization provider
2026-06-09Clawvisor Skill documentation published showing task-scoped access patterns
2026-06-06Third-party analysis notes Microsoft's Entra Agent ID as competing runtime-policy gap
2026-05-23Third-party summary of NSA agentic AI authorization advisory published
2026-05-16Clawvisor open-core Go gateway published on GitHub
2026-05-13Coverage notes Anthropic co-authored MCP authorization spec with Microsoft, Okta/Auth0
2026-05-08hiretop.com profile of Clawvisor and safe AI agents published
2026-05-07Clawvisor publicly launches via Y Combinator Spring 2026 (P26) batch
2026-04-01NSA publishes 'Careful Adoption of Agentic AI Services' advisory
2026-02-01NIST NCCoE publishes concept paper on OAuth 2.0/2.1 extensions for agent identity

Revision history

27 July 2026

First assessed. Protection under test. The text above is fixed as of this date and is not edited afterwards.

In the same band

Ploy Standout Playabl.ai