BOOTMUX

The description states that BOOTMUX is a physical-first bridge for AI assistance. It combines:

• An iPhone client (SwiftUI, CoreBluetooth)
• An ESP32-S3 device (BLE, USB HID)
• A target-side Companion written in Go
• Codex CLI integration
• Terminal session handling via PTY and WebSocket

The system enables two separate paths:

1. Input path: iPhone → BLE → ESP32-S3 → USB HID keyboard input to the target machine.
2. Output path: Target PTY / Codex → Companion → WebSocket → iPhone terminal output.

It is described as not being an SSH client, but rather a way to get AI assistance before a computer is ready for AI tools like Codex.

Evidence strength Self-reported and unverified.

The author positions BOOTMUX as solving a gap in AI usability: “AI is available during almost every part of my work—except the moment I need it most.” The system aims to bridge the physical first mile for AI tools like Codex, especially when systems are not yet configured or reachable.

Key claims:

• BOOTMUX creates the physical first mile for Codex.
• It works before SSH or network paths exist.
• It separates input and output to avoid false confirmation of processing.
• The system was built using an AI-assisted loop involving Codex and GPT-5.6.

Evidence strength Self-reported, unverified claims about intent and positioning.

The description does not explicitly name a target customer or define an Ideal Customer Profile (ICP). However, the author implies that BOOTMUX is for developers or users who need AI assistance during early setup or recovery of machines where standard tools like SSH are not yet available.

It may appeal to:

• Developers working in environments with limited infrastructure
• Users needing remote access to broken or unconfigured systems

Evidence strength Inferred from context; no explicit customer definition.

There is no evidence of a business model, pricing structure, or monetization strategy in the description. The project is presented as a Build Week hackathon submission with no indication of commercial intent or revenue streams.

Evidence strength Not evidenced.

The system uses:

• iPhone (SwiftUI, CoreBluetooth)
• ESP32-S3 (BLE, USB HID)
• Go for the Companion
• Codex CLI integration
• PTY and WebSocket for terminal handling
• ARM64 Lima VM for testing

It includes:

• Bounded input/output queues
• Session isolation
• UTF-8 handling
• Fail-closed lifecycle rules
• Streaming ANSI sanitization

The author states that this was built in a Build Week period using an AI-assisted loop involving Codex and GPT-5.6.

Evidence strength Self-reported, with no independent verification of technical claims or delivery.

There is no evidence of traction, customers, or adoption beyond the author’s own demonstration. The project is described as a Build Week hackathon submission and not yet production-ready.

The description states:

• Demonstrated paths include ASCII input and Codex output.
• Judge Mode allows inspection without rebuilding hardware.
• No repeatable production-ready physical operation has been achieved.
• Future work includes copy/clear acceptance, HID Mirror, and broader platform testing.

Evidence strength Not evidenced.

The description does not mention any competitors or market context. It is unclear whether similar systems exist or how BOOTMUX compares to existing solutions for remote terminal access or AI-assisted development in early setup scenarios.

Evidence strength Not evidenced.

• No commercial traction or revenue data: The project is described as a hackathon submission with no evidence of adoption.
• Unproven scalability or usability: The system is not yet production-ready and lacks full terminal emulation, Unicode support, or offline operation.
• AI-assisted development loop dependency: Reliance on Codex and GPT-5.6 may not be scalable or replicable outside the author’s specific setup.
• Limited hardware compatibility: Only ESP32-S3 is mentioned as supported, with no indication of broader platform support.

Evidence strength Inferred from self-reported limitations and lack of external validation.

1. What is the actual end-user problem you are solving, and how do you know it exists?
2. Have you validated this solution with real users or in real-world settings beyond the Build Week demo?
3. How does BOOTMUX handle edge cases like network disconnections or hardware failures?
4. Is there a plan to scale beyond a single developer’s use case?
5. What are the technical and legal implications of using Codex and GPT-5.6 for system-level development?
6. Are you planning any monetization strategy or commercial model?

Evidence strength Inferences based on self-reported description.