Refresh pipeline contract docs to match current runtime behavior

Inject architecture section only in accordion, keep docs file clean

Inject architecture diagram by heading and ignore runtime timing log

Fix docs marker placement and simplify Space README

Render architecture diagram via HTML at docs marker for HF compatibility

Render architecture diagram as embedded static PNG in docs section

Render Mermaid from flattened paragraph blocks in HF markdown

Revert Mermaid fallback changes and restore prior diagram renderer

Fix Mermaid fallback rendering and refresh README overview

Update docs diagrams, evaluation labels, and analysis artifacts

Improve user-facing Suggested Prompt error messages

Add corporate tag exclusions and increase row display defaults

Restore hover tooltips with robust row-index mapping

Remove probe sequential species experiment; keep probe pick cap

Tune probe selection budget and optional sequential species pass

Fix UI tag-button desync and add regression smoke coverage

Consolidate probe configs and eval artifacts on main

Add retrieval model artifacts for Spaces runtime

Add Stage 3 fast retry, stricter preflight, and timeout tuning

Stop executor shutdown from blocking after LLM timeouts

Add missing langchain dependencies for Stage 3 selection

Track PNG assets with Git LFS

Track wiki_pages CSV with Git LFS

Align Gradio dependency with HF Space SDK version

Refine UI guidance, tag toggle cues, and hover tooltip metadata

Polish Gradio UI layout/tooltips and remove dead helper code

Refine tag toggle UI ordering/colors and add category assignment analysis artifacts

Add synchronized lego-style tag toggles and prompt builder UI

Speed up retrieval exact/alias fasttext path and rerun n30 eval

Add n30 caption-evident set and per-group display blend overrides

Switch Stage3 to explicit-only no-why selection, drop bear probe, and set k=1 defaults

Consolidate pending pipeline, structural, and analysis updates

Remove dead retrieval/display helpers and simplify debug paths

Remove Stage3 bootstrap rerank and add experiment memory log

Simplify Stage3 chunking to interleave-only and add eval diagnostics

Add eval audit tools, caption-evident set, and logging

Record non-fatal pipeline issues in eval JSONL outputs

Clarify stage3_debug rewrite-first behavior in quickstart docs

Remove outdated branch list from project summary

Add mandatory workspace bootstrap steps for Git alignment

Add comprehensive project documentation for session continuity

Fix eval_categorized.py to work with eval_pipeline.py output

Remove binary PNG files (use Hugging Face XET storage instead)

Add eval results for debugging

Add ranking metrics infrastructure to eval pipeline

Add per-category evaluation metrics script

Fix constraints: body_type, species, gender are multi-select

Add tag categorization pipeline for e621 checklist

Improve topless/bottomless definitions to prevent confusion

Fix Windows encoding issues in diagnostic script