White paper ยท 2026-06-07

Sheriff Lone Star: from household report to researched resolution.

A quick operating paper for the Las Jaras issue desk: how the public form captures household problems, how Sheriff Lone Star triages severity, and how chat-based research turns a report plus media into practical next steps.

Executive Summary

The Sheriff Lone Star form is a lightweight, public, ADA-minded issue-intake desk for Las Jaras. Its job is to turn messy household observations into trackable cases with enough context for a useful investigation: reporter, room, issue type, urgency, photo or video evidence, desired outcome, and possible action lanes.

The system does not pretend the static page can inspect uploaded media by itself. It saves local file metadata and previews for the human, then generates a chat handoff prompt. The human uploads the same photo or video into chat, pastes the prompt, and Sheriff Lone Star researches the issue with the media and report context together.

For production use, the Sheriff board needs to move beyond browser-local storage. The prototype could open cases quickly, but local storage belongs to one browser on one device, which means a case entered on a phone may not appear on a laptop, in an admin view, or in a later resolution pass. Cloudflare Worker plus D1 gives Sheriff Lone Star a real shared case API: the static Labs page can create, list, update, check off, and close cases across devices without exposing private write credentials in the public HTML.

Design Principles

Issue Resolution Flow

1. Intake Capture who reported it, where it happened, what changed, the category, urgency, desired outcome, evidence description, and photo or video metadata.
2. Triage Assign severity, protect people and property first, and identify whether this is research, DIY, shopping, vendor, maintenance, or conversation work.
3. Investigate Use visible evidence, Las Jaras context, follow-up questions, and grounded research to separate likely causes from guesses.
4. Resolve Recommend the next action, track owner and target timing, suggest supplies or pros when relevant, and close only when the human confirms Case Closed.

Severity Log

Severity Use When Response Target
S1 Emergency Possible life-safety, fire, gas, shock, flooding, sewage, injury, break-in, unsafe access, or immediate property-damage risk. Stabilize now; investigate immediately. Stop damage before researching elegantly.
S2 High Plumbing trouble, no hot water, active leak, no water, broken lock, HVAC in severe weather, pest concern, mold, or same-day household disruption. Act today or within 24 hours. Decide DIY versus professional help quickly.
S3 Standard Running toilet, slow drain, broken fixture, plant mystery, household interpersonal friction, shortage, repair, or important non-emergency issue. Research and schedule within 2 to 7 days. Bundle related work when possible.
S4 Watch Low-risk oddities, cosmetic issues, uncertain spiritual heartbeat notes, or cases that need more evidence before spending money. Monitor, batch, or revisit within 1 to 2 weeks.

Research And Prompt Workflow

  1. The form creates a browser-local case and a first read from Sheriff Lone Star.
  2. The form generates an investigation prompt with case title, reporter, room, category, urgency, severity, desired outcome, action lanes, media metadata, evidence description, and issue description.
  3. The human uploads the same photo or video into chat and pastes the prompt.
  4. In chat, Sheriff Lone Star examines the attached media, uses the report context, and researches likely causes when outside information would improve accuracy.
  5. The answer is separated into visible evidence, inferred causes, checks performed, likely explanation, follow-up questions, solution options, and next action.
  6. If the path is clear, Sheriff Lone Star recommends DIY steps, a TaskRabbit/pro option, H-E-B list items, a maintenance task, or a household conversation script.

Core boundary: the form can preview media and save file metadata, but the actual visual analysis happens in chat after the human attaches the media there.

Best-Practice Sources Folded Into The Flow

Artifacts

Open Sheriff Lone Star Issue Desk Open Serafina Field Log Cloudflare Backend Build Log