Hubble

The company appears to be a solo developer project submitted to the OpenAI 2026 hackathon. The author describes Hubble as an AI agent coordination tool that allows users to delegate tasks to AI agents (referred to as "Codex") while maintaining control over what information is shared and when connections are made. The system operates through a macOS desktop app with a focus on permissioned interactions between human users and AI agents.

The description states that Hubble is "a desktop-first macOS experience where a user gives their Codex a pet identity and asks, 'Find two beta testers for my app.'" It describes a system where:

• A user requests a task through a Codex agent
• This request becomes visible in "Hubble Commons"
• Agents discover the request and respond
• Every meaningful step stops at an owner-controlled disclosure or approval gate
• When both sides approve, Hubble opens a mission-only social Room and writes a Receipt

The demo includes one real, bounded Codex CLI step that calls an isolated local Hubble MCP, reads a fixed public-safe capsule, and submits structured interest for a candidate. However, the description explicitly states that "Interest is not consent. Counterpart approvals, the second match, and final results are explicitly labeled local demo fixtures."

The system uses SwiftUI for the native shell, with local Three.js scenes running inside WKWebView using Blender-authored GLB assets. Native Swift owns mission state, approvals, Rooms, Records, and deterministic local Receipts.

Not evidenced No actual functionality beyond demo fixtures is demonstrated or claimed to exist.

The author states: "Powerful AI can write, build, and learn, but it still lives inside a chat box. Hubble asks: what if your AI could go outside for you—to find a friend, customer, beta tester, or collaborator—while you remain in control?"

This positions Hubble as an extension of AI agents beyond traditional chat interfaces, focusing on external coordination and collaboration while maintaining user control.

The project also states: "An agentic social product is not credible because agents look autonomous. It becomes credible when intent, disclosure, approval, and evidence are visible."

Inference The positioning suggests a shift from purely conversational AI to collaborative AI systems with explicit permission flows and transparency mechanisms.

The description states that Hubble allows users to ask Codex to "find two beta testers for my app." It also mentions that the system could be used to find "a friend, customer, or collaborator."

Not evidenced No specific target customer segments are named or described beyond general use cases like beta testing, finding friends, customers, or collaborators.

The description does not provide any information about pricing, monetization strategies, or business model. It only describes the conceptual framework and demo functionality.

Not evidenced No evidence of a business model or pricing structure is provided.

The project was built with SwiftUI and uses local Three.js scenes inside WKWebView with Blender-authored GLB assets. Native Swift handles mission state, approvals, Rooms, Records, and deterministic local Receipts.

The Codex experiment uses a narrow local MCP boundary with no access to HOME memory, private contacts, repository data, payment authority, or live backends.

Inference The technical approach suggests a focus on local execution and controlled boundaries for AI interactions, emphasizing security and user control.

The description explicitly states that this is a hackathon submission to the OpenAI 2026 hackathon. It includes a demo that performs no real outreach, messaging, scheduling, payment, or identity disclosure.

The author notes: "The demo includes one real, bounded Codex CLI step... Interest is not consent. Counterpart approvals, the second match, and final results are explicitly labeled local demo fixtures."

Not evidenced No traction, revenue, customers, or adoption data beyond this self-reported hackathon project is available.

No competitive analysis or market positioning is provided in the description. The author does not reference existing products or services that might compete with Hubble.

Not evidenced No information about competitors or the broader marketplace context is included.

• Demo-only functionality: The system performs no real outreach or disclosure, and all interactions are demo fixtures.
• Solo developer project: Only one team member (D2 R2) is listed, suggesting limited development capacity.
• No commercial traction: No evidence of revenue, customers, or adoption beyond the hackathon submission.
• Unproven concept: The described functionality has not been implemented in a real-world setting.

1. What specific problems are you trying to solve with Hubble?
2. How do you plan to transition from demo fixtures to actual AI agent coordination?
3. What are the key technical challenges in implementing the permissioned interaction model?
4. Have you identified any potential regulatory or compliance issues related to AI agent interactions?
5. What is your roadmap for scaling beyond the current hackathon prototype?