fix(agent): disable response buffering for chat streaming

fix(agent): forbid re-pasting tool output in the text reply

feat(embed-studio): make agent-generated charts theme-aware

fix(agent): default back to gpt-oss-120b, pin Llama 3.3 to Together

fix(agent): strip reasoning_content from history, default to Llama 3.3 70B

feat(agent): switch AI backend from OpenRouter to HF Inference Providers

feat(editor): embed studio with data files and agent-aware editing

refactor(backend): modular server split with new routes, persistence and agent layer

chore: initial commit