fix: eliminate duplicate events and AttributeError crash in demo runner

fix: detect interrupt() via Interrupt objects, not current.next

fix: remove unused REPAIR_REVIEW import in demo_runner (F401)

fix: remove duplicate REPAIR_REVIEW render in demo_runner apply_event

docs: audit, HF Spaces fix, and deep inline walkthrough comments

fix: resolve ruff errors in demo/app.py (F821, F841 x2)

feat: demo enhancements — per-role routing, HITL end-to-end, cleaner timeline

fix: revert max_new_tokens, improve error display, forbid comments in generated code

fix: increase token budget for long tasks and show all 4 example buttons

feat: replace example tasks with harder problems to trigger repair loop

fix: guard event stream against non-dict items from parallel reducer

feat: switch to Claude API (Haiku 4.5), production-only mode, parallel spec+solution

fix: replace PNG graph images with inline SVG to fix HF sync

chore: track PNG files with git-lfs for HuggingFace

feat: add HuggingFace persistent cross-session memory (hf_memory_sync)

feat: replace Graph Topology tab with static PNG images

chore: fix lint warnings (ruff --fix)

Phase 4 upgrades: fixes 10, 16, 18, 19

Phase 3 upgrades: fixes 6, 8, 14, 15, 17

Phase 2 upgrades: fixes 5, 7, 9, 12, 13

Phase 1 upgrades: fixes 1, 2, 3, 4, 11

docs: add Demo screenshot, Development Notes attribution, and deployment knowledge base

docs: rewrite README as concise recruiter-ready portfolio artifact

Fix JSON parse failure: raise token limits and add truncation salvage

Upgrade Gradio 5.9.1 → 6.6.0 to fix Python 3.13 lock initialization bug

Fix Gradio 5 queue NoneType crash: call demo.queue() before launch()

Fix WebSocket connection error: bind server to 0.0.0.0 for HF Spaces proxy

Fix two remaining Python 3.13 / Gradio 5 incompatibilities

Fix Gradio/pydub crash on Python 3.13: pin gradio==5.9.1, replace BarPlot

Fix Python 3.13 crash: upgrade Gradio 4 → 5 (audioop removed in 3.13)

Fix Python 3.13 threading crash on HF Spaces startup

Fix README metadata: shorten short_description and correct sdk_version

Fix: move sync workflow to .github/workflows so GitHub Actions picks it up

Revise README with new project details and structure

Modify README metadata for Self Healing Code Agent

Add GitHub Actions workflow to sync to Hugging Face

Add Colab notebook, HF Spaces model pre-warming, and fix HF provider bugs

Fix QA JSON parse failure and Gradio event loop crash

Fix schema validation failures: tighten prompts and add output coercion

Fix Ollama ReadTimeout: raise inference limit to 600s for CPU inference

Fix Ollama auto-detection: replace async event loop health-check with sync

Implement complete Self-Healing Code Agent framework