Add LLM agent guidance files and player component architecture docs

[Copilot]

No guidance exists for LLM agents working in this repo, and there is no design documentation for the planned player component. Establishes the convention and content for both.

New files

• AGENTS.md — Root-level agent guidance: component inventory, architecture summary, shell script conventions, key design decisions (UDP multicast, MPEG-TS, latency-first)
• .github/copilot-instructions.md — GitHub Copilot auto-loaded instructions; surfaces critical context (raw UDP multicast, not HLS/DASH) and cross-references the other docs
• docs/architecture.md — Full system design: transport diagram, stream composition (H.264 + AC3 PIDs), timing/latency parameter table with rationale
• docs/player-architecture.md — Player design spec: goals, chosen approach (FFplay wrapper shell script), alternatives considered and rejected, required FFplay flags, configurable variables, file layout, explicit out-of-scope items
• player/AGENTS.md — Player-scoped agent guidance: four hard constraints (no extra latency, no new deps, no signalling, variables not literals), files to create, manual test procedure

Conventions established

All guidance files follow a consistent pattern: each significant component directory gets an AGENTS.md that states design intent, constraints, and how an agent should approach modifications — rather than leaving that context implicit.

Original prompt

I'm working on a new player component. There isn't a repo for it yet. I want to establish agents.md files (or similar) that provide guidance to the LLM, and high level design guidance to help the LLM understand the architecture. What's the best way to do this?

---

📍 Connect Copilot coding agent with Jira, Azure Boards or Linear to delegate work to Copilot in one click without leaving your project management tool.