docs: add 7-slide hackathon pitch deck (pptx + pdf)

docs: embed demo video in README · live 5-agent pipeline run

docs: add 5 dashboard + benchmark screenshots

docs: wire 5 dashboard/benchmark screenshots into README

docs: hackathon-ready README — V6.0 metrics, Mermaid arch, TAM/SAM, INV-15

fix: write benchmark_v5_results.json to demo/ (relative path)

fix: test_mcp_server 12 failures resolved — model fields, registry API, GPU label

fix: sync test_mcp_server imports with current server/models API

fix: S04 status corrected to DONE, screenshots section updated, Engineering Principles count corrected to 10

feat: V6.0 — TokenDance Master-Mirror storage, JCR Safety Gate (INV-15), AITER ROCm config. 15/15 PASS

fix: S-3 rotate_kv_quantization 4D indexing, S-13 speculative acceptance rate, Gradio real pipeline data

docs: update README with real MI300X benchmark results and dashboard screenshots

fix: Gradio 6.0 compatibility - gr.Table→Dataframe, theme, return values, remove download

Merge branch 'main' of https://github.com/SuarezPM/ContextForge

feat: APOHARA: Context Forge V5 — synthesis + rebrand complete

Revise README formatting and content

README V5 creative rewrite — APOHARA brand, 8 mechanisms, ICML 2026 QueueingController, status table S01-S14, 14 invariants

README V5 complete + test_queueing_controller.py

ContextForge V4/V5: Embeddings module + dedup/registry cleanup

PBKVPredictor: 2nd-order Markov model, 19 tests (stub → production)

ContextForge V5.0 PREVIEW: QueueingController, VisualKVCache, SpeculativeCoordinator, PBKVPredictor Markov, Dashboard, DevCloud runner

ContextForge V4.0: Benchmark V4 + 9 test files

ContextForge V4.0: EmbeddingEngine, CLA metadata, RotateKV, step graph

ContextForge v3.0: production-grade shared context compiler

ContextForge v2.0: production-grade shared context compiler

ContextForge v0.1.0 - shared context compiler for multi-agent LLM systems