docker: production image + Gitea Actions for monthly refresh

Dockerfile: self-contained image with corpus + Chroma + BM25 baked
in. Drawbar's compose pulls + runs without volume mounts. Built from
sources.json (labels schema), PRODUCT_NAME=crop_chem by default,
HYBRID_SEARCH=true (always-on for production quality). RERANK_URL +
OLLAMA_URL get set at compose time.

.gitea/workflows/refresh.yml: monthly cron (1st @ 06:00 UTC) does
full scrape → reindex → image push. Scrapes Bayer (~30 min) +
EPA PPLS row-crop filtered (~7h). Skips reindex+push if no corpus
diff. Tags pushed: :latest, :<sha12>, :corpus-<YYYY.MM.DD>.

.gitea/workflows/image-only.yml: on-demand or auto on code-only
pushes to main (paths: docs_mcp/, rag/, scrape/, requirements.txt,
Dockerfile, sources.json). Reindexes from committed corpus, builds
image, pushes. ~10 min vs ~9h full refresh.

.gitignore: corpus/ now COMMITTED (4,159 labels, 265 MB of .md +
sidecars). Lets image-only.yml rebuild indexes without re-scraping.
chroma/ + bm25/ still gitignored (regenerable binary indexes).

.dockerignore: drops venv, eval results, PLAN/README/CLAUDE.md,
deploy/, .git/ — keeps the image lean. corpus + chroma + bm25
explicitly NOT in dockerignore (those go INTO the image).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

corpus/epa_ppls/34704-41.md

@@ -0,0 +1,61 @@
+# CLEAN CROP DIAZINON AG500 INSECTICIDE
+
+- EPA Reg No: **34704-41**
+- Registrant: LOVELAND PRODUCTS, INC.
+- Signal word: Caution
+- Active ingredients: Diazinon (48%)
+- Label accepted: 2004-01-06
+- Source PDF: https://www3.epa.gov/pesticides/chem_search/ppls/034704-00041-20040106.pdf
+
+---
+}/r r-A"7i"'inE ..... P::;,T=-=B=-=D="" 
+JAN - 6 2004 
+SUPPLEMENTAL LABEL 
+DIAZINON AG 500 
+EPA REG. NO. 34704-41 
+DIRECTIONS FOR USE 
+It Is a violation of Federal Law to use this proci~ct In a manner Inconsistent with Its labeling. 
+READ ENTIRE LABEL AND PAMPHLET, USE STRICTLY IN ACCORDANCE WITH PRECAUTIONARY 
+STATEMENTS AND DIRECTIONS AND WITH APPLICABLE STATE AND FEDERAL REGULATIONS. 
+FOR PROTECTION OF AQUATIC SPECIES IN COUNTIES AND PORTIONS OF 
+COUNTIES LOCATED IN THE SACRAMENTO AND SAN JOAQUIN VALLEYS 
+BELOW 1000 FEET ELEVATION FROM POTENTIAL CONTAMINATION IN 
+SURFACE WATERS OF DIAZINON APPLIED DURING THE DORMANT SPRAY 
+PERIOD, THE FOLLOWING ADDITIONAL LABEL REQUIREMENTS APPLY: 
+• Dormant applications on orchard crops are restricted to ground application equipment 
+only. 
+• Do not apply within 100 feet upslope of 'sensitive aquatic sites' such as any irrigation 
+ditch, drainage canal or body of water that may drain into a river or tributary unless a 
+suitable method is used to contain or divert runoff waters. Waters that are contained or' 
+diverted must be held for a minimum of 72 hours before release into a sensitive aquatic 
+site. 
+• Maintain a vegetative buffer strip a minimum of 10 feet wide from the edge of a field that 
+is adjacent to and within 100 feet of sensitive aquatic sites. 
+• Do not apply this product to orchards when soil moisture is at field capacity, and/or when 
+a storm event likely to produce runoff from the treated orchard is forecasted by 
+NOAAlNWS, (National Weather Service) to occur within 48 hours following application. 
+• Make dormant applications only when insect scouting information or the recommendation 
+of a Pest Control Advisor indicate treatment is required. (See UC IPM Guidelines for 
+San Jose Scale in stone fruits and almonds and aphids in stone fruits. Use the prune 
+dormant spur sampling program to determine need for a dormant treatment in that crop). 
+• Apply only when wind speed is 3- 10 mph at the application site as measured by an 
+anemometer outside of the orchard on the side nearest and upwind from a sensitive site. 
+• When sensitive aquatic Sites are downwind from orchards, spray the first three rows 
+nearest the sensitive aquatic sites only when the wind is blowing away from the Sites. 
+The row at the edge of the field next to sensitive aquatic sites must be sprayed with the 
+outside nozzles turned off. Spray must not be directed higher than the tree canopy and 
+spray must be directed away from sensitive aquatic sites. 
+• The Stewardship Bulletin "Orchard Practices for Protecting Surfece Water", must be 
+available to handlers and equipment operators at the application site during all : , • , : , 
+application activities. (note: Bulletin available through CURES) • . ~ . • • 
+THIS LABELING MUST BE IN THE POSSESSION OF THE USER AT THE TIME OF •• • 
+PESTICIDE APPLICTION, PLEASE REFER TO CONTAINER LABEL FOR ADDITIONAL 
+PRECAUTIONARY STATEMENTS. 
+• 
+Loveland Products Inc. 
+P.O. Box 1286 
+Greeley, CO 80632 
+• • • • • . , .. 
+• . ' 
+· . " . ~ . · ,. • • 
+· ....... . , .. · . ,