Add framework-agnostic GitHub Copilot seed prompt for Voice WebRTC AI Programmer scaffolding

[Copilot]

Issue #1 requires a ≤500-character prompt to bootstrap the Voice WebRTC AI Programmer repository—a cross-platform app enabling voice-driven code editing via OpenAI Realtime API, MCP, and GitHub Copilot.

Changes

SEED_PROMPT.md

• 482-character framework-agnostic seed prompt instructing Copilot to scaffold cross-platform app structure
• Defers framework decision to evaluation phase per Issue Create bespoke self-proclaimed DIY license #1 requirements
• Specifies directory layout: src/webrtc/, src/realtime/, src/mcp/, src/github/, src/ui/, tests/, .github/workflows/
• Defines core UI components: connection controls, mic toggle, logs viewer, repo selector
• Framework evaluation approach comparing Flutter, React Native, Kotlin Multiplatform, and Qt
• Evaluation criteria: WebRTC support, Realtime API integration, MCP client viability, cross-platform reach, developer experience, ecosystem maturity
• Includes test infrastructure and CI/CD foundation
• Usage instructions and architectural overview

copilot-seed-prompt.txt

• Plain text seed prompt for direct copy-paste into GitHub Copilot Chat

The Seed Prompt

Create cross-platform Voice WebRTC AI Programmer (Android/iOS/Desktop). Scaffold: src/webrtc/ (mic streaming), src/realtime/ (OpenAI Realtime API client), src/mcp/ (MCP protocol), src/github/ (Copilot integration), src/ui/ (connect, mic toggle, logs, repo selector), tests/, .github/workflows/. Add README.md, docs/FRAMEWORK_EVAL.md (compare Flutter/KMP/Qt/ReactNative: WebRTC, Realtime API, MCP, cross-platform, dev experience, ecosystem). Voice→Realtime→MCP→Copilot→code workflow.

Rationale

Issue #1 explicitly states "the choice is not predetermined" and requires framework evaluation as a major early milestone. The seed prompt intentionally defers the framework decision to allow proper evaluation of:

• Flutter: Strong WebRTC support, cross-platform reach, rich UI toolkit
• React Native: Lightweight Node.js ecosystem, extensive npm packages, web developer familiarity
• Kotlin Multiplatform: Native performance, shared business logic
• Qt: Mature framework, excellent desktop support

The evaluation will produce a data-driven framework recommendation rather than a premature selection.

Optimizations and Enhancements

Based on feedback, the prompt has been optimized for brevity, clarity, and value:

• Clarity: Expanded "DX" abbreviation to "dev experience" for better readability
• Brevity: Removed obvious standard files (.gitignore, LICENSE) as they're implicit
• Conciseness: Shortened FRAMEWORK_EVALUATION.md to FRAMEWORK_EVAL.md
• Quality: Added tests/ directory for test infrastructure (TDD, quality assurance)
• Automation: Added .github/workflows/ for CI/CD foundation (automated testing, linting, builds)
• Result: 482 characters with 18-character budget remaining

Establishes Voice→Realtime API→MCP→GitHub Copilot→code changes as the core data flow, with all components and quality practices represented in scaffolding.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>1️⃣ Project 1 — Voice WebRTC AI Programmer</issue_title>
<issue_description># 1️⃣ Project 1 — Voice WebRTC AI Programmer
Epic: Build the tool we will use to build everything else.

This project creates a cross-platform Voice WebRTC AI Programming Assistant that lets us speak to write, edit, and manage code across GitHub repositories using OpenAI Realtime API + MCP + GitHub Copilot.
It is the foundational tool for the entire X-project series #4.

---

🎯 Goal

Create an app (Android first, eventually iOS + Desktop) that allows:

• Speaking to generate/modify code
• Realtime WebRTC → OpenAI Realtime API → MCP → GitHub Copilot
• Automated file creation, editing, commits, and PRs
• A simple interface to view logs, diffs, and actions

This becomes the core AI-powered coding environment for later projects.

---

🛠️ Very First Step (Before Anything Else)

The initial task of this project is:

Task 0: Generate a ≤500-character GitHub Copilot prompt that creates the initial working repository.

This short prompt will instruct Copilot to scaffold:

• A minimal cross-platform app structure (framework TBD)
• Barebones folders for WebRTC, Realtime API, and MCP integration
• A placeholder UI and project metadata
• A starting point for subsequent issues

Once the repo is generated through this tiny seed prompt:

Next: Use GitHub Issues to evolve the project

Issues will then drive:

• Framework evaluation
• Platform architecture
• Feature expansion
• UI/UX evolution
• Integration details

The entire project grows from this initial seed prompt.

---

🛠️ Cross-Platform Framework Decision

This app must use a cross-platform framework, but the choice is not predetermined.

A major early milestone is:

Task: Evaluate several frameworks and recommend one.

Candidates:

• Flutter
• Kotlin Multiplatform
• Qt
• React Native
• Native+shared-core hybrid

Evaluation criteria:

• WebRTC support
• Realtime API integration
• MCP client viability
• Android/iOS/Desktop reach
• Developer experience & maintainability
• Ecosystem maturity
• Long-term viability

Deliverable: written recommendation + repo adaptation plan.

---

📡 Core Features (MVP)

• WebRTC microphone streaming
• Connect to OpenAI Realtime API
• Speech recognition
• AI → MCP tool invocation
• MCP → GitHub Copilot → code changes
• Basic UI: connect, mic, logs, repo selector
• View diffs and PR results
• Configurable intents

---

🚀 Stretch Features

• Code navigation inside the app
• Continuous dictation mode
• Push-to-talk vs. hands-free
• Audio output responses
• Multi-repo switching
• Desktop client support

---

📦 Deliverables

• A working Voice WebRTC AI programming assistant
• Initial repo created entirely by the ≤500-character Copilot seed prompt
• Sequential GitHub Issues for structured development
• Framework decision + justification
• Clean MVP implementation with documentation

---

🔄 Dependencies

• None — this is the first project.
• Produces the tool required for Projects Migrate project tracking from ChatGPT to GitHub Issues #2–Build a browser extension to redirect search engines to non-AI/web answers #7.

---

🧪 Definition of Done

• Repo successfully generated from the 500-character prompt
• Voice → Realtime → MCP → GitHub Copilot loop operates end-to-end
• At least one file can be created/edited/committed via voice
• PR creation works
• Documentation describes architecture + chosen framework

---

📌 Next Steps

• Write the ≤500-character Copilot seed prompt
• Create the initial repo using that prompt
• Open Issues to expand the scaffolding
• Begin framework evaluation
• Implement minimal MVP
• Test full “speak → code → commit → PR” workflow

This project unlocks the entire remaining X projects in the series #4.</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes 1️⃣ Project 1 — Voice WebRTC AI Programmer #5

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[paulpv]

@copilot Why not ReactNative? I would think the benefit of an arguably much lighter [programming?] weight Node.js framework would be overall more beneficial than using Flutter.

Does Flutter really provide so good of a "WebRTC support, cross-platform reach, and MCP client viability" that it is the obvious winner?

[Copilot]

@copilot Why not ReactNative? I would think the benefit of an arguably much lighter [programming?] weight Node.js framework would be overall more beneficial than using Flutter.

Does Flutter really ...

You're absolutely right. Issue #1 explicitly requires framework evaluation as a major early milestone, stating "the choice is not predetermined." I've updated the seed prompt to be framework-agnostic (commit 5e7d78e).

The new 477-character prompt now scaffolds docs/FRAMEWORK_EVALUATION.md to compare all candidates:

• React Native: Lightweight Node.js, extensive npm ecosystem, react-native-webrtc
• Flutter: Strong WebRTC support (flutter_webrtc), cross-platform reach
• Kotlin Multiplatform: Native performance, shared business logic
• Qt: Mature C++, excellent desktop support

Evaluation criteria include WebRTC support, API integration capabilities, MCP client viability, cross-platform reach, developer experience, and ecosystem maturity. The framework selection will be data-driven rather than predetermined.

[paulpv]

@copilot What is "DX"?

[Copilot]

@copilot What is "DX"?

"DX" stands for "Developer Experience." I've replaced it with "dev experience" for clarity (commit ab9f201). The prompt is now 489 characters, still within the 500-character limit.

[paulpv]

@copilot For brevity, you can leave out the obvious "Include .gitignore, LICENSE", and shorten FRAMEWORK_EVALUATION to FRAMEWORK_EVAL.

[Copilot]

@copilot For brevity, you can leave out the obvious "Include .gitignore, LICENSE", and shorten FRAMEWORK_EVALUATION to FRAMEWORK_EVAL.

Good suggestions! I've removed the obvious .gitignore and LICENSE mentions and shortened FRAMEWORK_EVALUATION to FRAMEWORK_EVAL (commit d1be20e). The prompt is now 454 characters, saving 35 characters and leaving 46 characters of budget available.

[paulpv]

@copilot Anything else useful/valuable you can add in the prompt to fill in those extra available characters?

[Copilot]

@copilot Anything else useful/valuable you can add in the prompt to fill in those extra available characters?

Added tests/ and .github/workflows/ to the scaffold (commit 56227da). These provide essential test infrastructure and CI/CD automation foundation. The prompt is now 482 characters with 18 characters remaining budget. This ensures quality practices are built in from the start.

[paulpv]

Repo created with prompt and populating with LookAtWhatAiCanDo/Codeoba#1