farmos-observations
by brianppetty
DevOps & CloudBrief
Query and create field observations and AI-processed captures.
Description
What This Skill Does
Logs and retrieves field observations in a self-hosted FarmOS instance. Handles pest, disease, weed, crop condition, weather damage, and equipment issue reports, with optional photo attachments. AI processing assigns urgency scores and extracts entities from free-text descriptions.
Accepts sparse natural-language field reports and auto-detects observation type, severity, and field location before creating structured records, removing the need for crew to fill out forms.
When to Use It
- Logging a weed pressure report from a field crew chat message
- Reviewing all high-urgency observations in a specific field this week
- Attaching a field photo to a new pest sighting for species identification
- Checking whether a recurring issue has prior observations before scheduling a spray
- Escalating a high-severity crop damage report to the operator immediately
Example Workflow
User asks: Found waterhemp in field 22, northeast corner near the waterway
- 1
Detects observation type as 'weed' and field ID 22 from the message
- 2
Identifies location as 'northeast corner near the waterway' and defaults severity to medium
- 3
Authenticates using the local farmos-auth.sh script to obtain a JWT token
- 4
Creates the observation via POST /api/observations with all extracted fields
- 5
Offers to create a follow-up work order and checks for similar weed reports in adjacent fields
Observation created for waterhemp in field 22, northeast corner, medium severity, with offer to open a spot-treatment work order.
Installation
clawhub install brianppetty/farmos-observationsRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.