Bitty — Codex That Runs When You Arrive

The description states that Bitty is:

• A local-first mobile app for iOS (iPhone/iPad)
• Designed to connect to a private Node.js runner on macOS
• Built with Expo, React Native, and TypeScript
• Uses Codex CLI integration, exposing session state, streaming responses, approvals, files, Git changes, and job information
• Communicates via authenticated HTTP and WebSocket connections
• Supports QR pairing and secure local storage
• Does not require a separate OpenAI API key — uses existing Codex authentication

The app enables:

• Starting, resuming, and monitoring Codex sessions
• Sending prompts by text or voice
• Listening to assistant responses with TTS
• Reviewing tool approval requests
• Inspecting Git diffs, workspace files, and running jobs
• Tracking token usage, context usage, and runtime status
• Managing multiple coding sessions through the Mini Board
• Receiving notifications when work completes or requires approval
• Triggering Codex runs based on location and time window

Inferred: The app is a companion tool for developers using local Codex workflows.

The author states:

• Bitty is intended as a local-first iPhone companion for Codex sessions
• It aims to make it easier to supervise local coding agents while moving around
• It does not aim to replace the desktop development environment

Inferred: The positioning reflects a niche use case — supporting developers who work locally but want mobile access and control.

The claim evolution shows:

• Initial version supported basic Codex session management
• During Build Week, it added location-triggered workflows, background recovery, and push notifications
• The author emphasizes that this is not about replacing desktop tools, but enhancing them

Not evidenced: No claims about market positioning, competitive differentiation, or strategic direction beyond the developer’s own use case.

The description states:

• Bitty targets developers using local Codex workflows
• It supports those who want to monitor and interact with Codex sessions from mobile devices
• The tool is designed for users who are already familiar with Codex CLI, Node.js, and macOS development

Inferred: The ICP likely includes:

• Developers working on local AI coding agents
• Users of Codex or similar tools (e.g., ChatGPT plugins, GPT-4)
• Tech professionals who value local-first and secure workflows

Not evidenced: No explicit customer personas, segmentation data, or user feedback.

The description states:

• Bitty uses the developer’s existing Codex authentication
• It does not require a separate OpenAI API key
• The app is built as a personal tool during a hackathon
• No pricing model, monetization strategy or revenue streams are mentioned

Inferred: There is no evidence of any business model or pricing structure.

Not evidenced: No information on monetization, subscriptions, licensing, or paid features.

The description states:

• Built with Expo, React Native, TypeScript
• Uses Node.js as backend runner
• Communicates via HTTP and WebSocket connections
• Integrates with Codex CLI
• Supports QR pairing, secure local storage, TTS, and approval handling
• Includes push notifications, background recovery, and multi-session support

Inferred:

• The tool is a developer-focused prototype built for personal use
• It uses modern mobile and backend stacks (React Native, Node.js)
• It handles complex technical challenges like reconnection logic, secure state sharing, and approval boundaries

Not evidenced: No evidence of production deployment, scalability, or performance metrics.

The description states:

• The app was built during a Build Week hackathon
• It supports features like location-triggered runs and background recovery
• The author reports improvements made during the hackathon
• No mention of users, adoption, or usage beyond personal use

Inferred: This is a proof-of-concept prototype, not a product with traction.

Not evidenced:

• No user base
• No customer feedback
• No revenue or monetization data
• No production deployment or long-term usage metrics

The description states:

• Bitty is a companion tool for developers using local Codex workflows
• It integrates with Codex CLI, which is used in AI coding environments
• The author does not reference any competitors directly

Inferred: The competitive context includes:

• Other tools that enable local AI agent workflows
• Tools that support mobile access to development tasks
• Platforms like ChatGPT plugins, local LLMs, or AI coding assistants

Not evidenced: No mention of existing tools, market analysis, or competitive positioning.

The description states:

• Bitty is a personal project built during a hackathon
• It uses developer-specific tools (Codex CLI, Node.js)
• The author is the only team member (1 person)
• No evidence of production use or scalability

Inferred risks and red flags:

• No commercial traction or user base
• Single-person development limits scalability and product maturity
• Limited market validation — no external feedback or adoption
• Highly niche use case — likely only useful to developers using Codex locally
• Unproven business model — no monetization strategy

Not evidenced:

• No risk analysis, financials, or user retention data
• No evidence of product-market fit or long-term viability

1. What is the actual use case for Bitty beyond your own development workflow?
2. Have you tested it with other developers or users outside of yourself?
3. Are there any plans to support more than one Codex authentication profile?
4. How do you plan to scale this beyond a single developer’s personal tool?
5. What are the technical limitations of running Codex workflows locally and remotely?
6. Is there any intention to monetize Bitty or offer it as a service?
7. How does Bitty handle edge cases like network interruptions, approval timeouts, or file access failures?