Migrate to GPT-5 only, simplify to -y flag, remove LiteLLM and extras

[shouryamaanjain]

Title: Migrate Emplode to GPT-5 only, remove LiteLLM and extras, and simplify CLI (-y only)

Summary

• Standardize on GPT-5 as the sole model; remove all alternative backends and flags.
• Replace LiteLLM with the official OpenAI client (v1.x); add non‑stream fallback when org streaming is restricted.
• Minimize CLI to only -y/--yes; remove fast/local/falcon/azure/model/api_base and all non‑core prompts.
• Strip external calls/prompts and marketing links; keep only core chat + run_code flow.
• Update dependencies to latest: openai 1.106.1, rich 14.1.0, tiktoken 0.11.0, tokentrim 0.1.13.
• Adjust runtime to temperature=1.0 for GPT‑5 compatibility and to end after executing code to avoid unsupported function role recursion.

Notes

• README updated to reflect minimal usage and single flag.
• Built and smoke‑tested: CLI prompt executes a code block that prints TEST_OK.

Impact

• Smaller surface area, fewer dependencies, and a simpler user experience.
• Single configuration: requires only OPENAI_API_KEY.

₍ᐢ•(ܫ)•ᐢ₎ Generated by Capy (view task)