ShadeLoop: an open-model field journal that gets you outside for 20 minutes

Description and use

ShadeLoop turns a 20‑minute outdoor request into a vetted mission, guides the user, records observations, and shares evidence for future users.

Problem

People want a quick, safe outdoor break but lack a structured plan that fits a 20‑minute window and records observations for future users.

Outcome

ShadeLoop transforms a 20‑minute outdoor request into a validated mission, guides the user, records observations, and shares evidence for future users, all within an offline‑capable PWA.

Workflow

• 1. User enters a mission description and allows location
• 2. Open-weight model (Gemma 4) parses request and drafts missions
• 3. Deterministic rule engine validates drafts against time, walking limits, accessibility, and past reports
• 4. Field Mode guides the user with read‑aloud steps, observation taps, and dictated notes
• 5. Observations are stored in Tiger Data (TimescaleDB + pgvector) and used to generate a report and similarity patterns
• 6. User can delete all data via Settings → Delete my data

Evidence

ShadeLoop finds you a shady bench, walks you there, and turns what you notice into evidence for the next person.

Source

It can shorten or refuse a draft, and then template missions are used and labelled as such.

Source

The same Discover flow as an installable, offline‑capable PWA

Source