Reptile Pet Health Diagnosis Tool

Problem

When keeping lizards, snakes, spiders, or other reptiles, visible signs such as scale shedding, skin redness, posture, and appetite can be hard to interpret without veterinary training. This skill is aimed at that gap: users provide a local video file or a public video URL, and the cloud service returns a structured reptile health assessment with follow-up report history.

How it works

The workflow is built around the scripts/crawl_analysis entrypoint and can be triggered inside an agent session when reptile pet analysis is requested. The main capabilities include:

• Video input: supports mp4, avi, and mov, with a 10MB limit; use --input for local paths and --url for remote URLs, which the API service downloads automatically.
• Visual analysis: examines scale condition, skin lesions, body appearance, and other visual markers, then provides common disease warnings and care suggestions.
• Species parameter: --crawl-type accepts values such as lizard, snake, spider, turtle, gecko, chameleon, scorpion, iguana, crocodile, or other.
• Historical reports: --list queries the cloud API only, avoids local memory or manual aggregation, and renders a Markdown table with report links.

A typical run involves preparing a clear video input, invoking the skill script, interpreting the structured result, and checking historical reports when needed. User identity is handled internally and should not be displayed or requested as a user parameter.

Boundaries

This tool is useful for everyday health reference and early warning clues, but it does not replace a professional veterinary diagnosis. Videos should clearly show the pet's body surface and fit the reptile pet scenario. Only the bundled skill scripts should be used; do not generate temporary replacement scripts. Historical report queries must rely on the cloud API response.