{
"event": "PreToolUse",
"tool_name": "Bash",
"tool_input": {
"command": "cd \/var\/www\/scripts\/pipeline && DB_PASSWORD=\"v8mFLmkF2fth_r\" .\/venv\/bin\/python -c \"\nfrom qdrant_client import QdrantClient\nimport requests\n\nclient = QdrantClient(host='localhost', port=6333)\n\n# Hole ein Embedding von Ollama\nresp = requests.post('http:\/\/localhost:11434\/api\/embeddings', json={\n 'model': 'mxbai-embed-large',\n 'prompt': 'Was ist systemisches Coaching?'\n})\nvector = resp.json()['embedding']\nprint(f'Embedding: {len(vector)} dims')\n\n# Suche in Qdrant\nresults = client.search(\n collection_name='documents',\n query_vector=vector,\n limit=5,\n with_payload=True\n)\n\nprint(f'\\\\n=== Ergebnisse ({len(results)}) ===')\nfor i, r in enumerate(results):\n title = r.payload.get('document_title', 'N\/A')[:40]\n content = r.payload.get('content', '')[:80].replace('\\n', ' ')\n print(f'{i+1}. Score: {r.score:.4f} | {title}')\n print(f' {content}...')\n\"",
"timeout": 60000,
"description": "Direct Python Qdrant test"
}
}