Identify plants, detect diseases, and provide biodynamic care recommendations using Pl@ntNet API. Supports image URLs and local files. Returns species name, confidence score, care tips aligned with lunar calendar, and treatment recommendations.
E*NKI uses this skill to identify plants from images, diagnose potential issues, and offer biodynamic care recommendations aligned with the lunar calendar.
{baseDir}/scripts/identify.py <image_url_or_path> [organ] [lang]
organ (optional): leaf, flower, fruit, bark, auto (default: auto)
(optional): , , , , (default: )
frenesdeptfrThe script uses its own venv with requests installed.
From image URL:
python3 {baseDir}/scripts/identify.py "https://example.com/plant.jpg" leaf
From local file:
python3 {baseDir}/scripts/identify.py "/path/to/photo.jpg" auto
The script returns a JSON object:
{
"species": "Solanum lycopersicum",
"common_name": "Tomato",
"confidence": 94.2,
"family": "Solanaceae",
"gbif_link": "https://www.gbif.org/species/...",
"edible": true,
"toxic": false,
"top_matches": [
{"species": "Solanum lycopersicum", "score": 0.942},
{"species": "Solanum melongena", "score": 0.041}
],
"enki_advice": "Tomate détectée. Plante racine selon le calendrier biodynamique..."
}
When this skill returns a result:
toxic: true — always warn the userAPI_KEY_MISSING → Tell user to set PLANTNET_API_KEY environment variable (free at identify.plantnet.org)IMAGE_UNREACHABLE → Ask user to send the image directly or verify the URLNO_MATCH → Apologize, suggest retaking photo in better light or different angleRATE_LIMIT → Pl@ntNet free tier is 500 req/day. If hit, notify user and retry tomorrowhttps://my-api.plantnet.org/v2/identify/all