Export PDF text chunks from Qdrant to a plain text file.
npx skills add https://github.com/techybolek/child --skill qdrant-file-exporter
Please help me install this Agent Skill. Skill: qdrant-file-exporter Source: https://github.com/techybolek/child/tree/main/.claude/skills/qdrant-file-exporter Command: npx skills add https://github.com/techybolek/child --skill qdrant-file-exporter
This Skill enables exporting all text chunks for a specific PDF document stored in the Qdrant vector database to a plain text file, including the three-tier contextual embeddings (master context, document context, and chunk-specific contexts), for offline inspection, analysis, or auditing.
Use the qdrant file exporter to dump all chunks for a specific PDF from Qdrant into a text file:
High-intent search queries and answers about installing and using this skill.
Export all text chunks by running the qdrant-file-exporter script with your PDF filename. The tool connects to Qdrant, filters chunks by filename, retrieves them in batches, sorts by chunk index, and saves to a plain text file in the UTIL directory with full contextual metadata.
Exported chunks include three-tier contexts: master context, document context, and chunk-specific context. This hierarchical metadata preserves the semantic relationships and enhances offline inspection and analysis of each extracted chunk.
Yes, chunks are automatically sorted by chunk_index during export, maintaining the original document order. This ensures sequential fidelity and readability in the output text file.
The exporter handles documents at scale by scrolling through results in batches of up to 100 chunks per request. This batch processing enables efficient retrieval and export of PDFs with hundreds or thousands of chunks.
Configure QDRANT_API_URL and QDRANT_API_KEY environment variables to connect to your Qdrant instance. Ensure the qdrant-client dependency is installed and your PDF has been indexed into the tro-child-3-contextual collection.
Yes, the plain text output is designed for offline inspection, compliance auditing, and detailed analysis of document chunks outside the vector database, with all contextual layers preserved for comprehensive review.
techybolekReload a single PDF's chunks into Qdrant without touching other documents.
GongLingRuiSplit long text into uniform chunks at sentence or paragraph boundaries.
DexploarerDeploy Qdrant vector databases on Railway with collection management and vector operations.
HCMUTE-RTICCapture, serialize, and retrieve project context in JSON, Markdown, or vector databases.
