Convert local documents to Markdown, chunk content, and ingest into Milvus.
npx skills add https://github.com/mengsi16/brain-base --skill upload-ingest
Please help me install this Agent Skill. Skill: upload-ingest Source: https://github.com/mengsi16/brain-base/tree/main/skills/upload-ingest Command: npx skills add https://github.com/mengsi16/brain-base --skill upload-ingest
When upload-agent receives a local document ingestion request, this workflow triggers the doc-converter for format conversion and dispatches the results to knowledge-persistence for chunking and Milvus ingestion. It runs in parallel with get-info-workflow and avoids external补库 chains.
Provide local document paths to ingest and the system will validate, convert formats, generate frontmatter, chunk, and store results.
High-intent search queries and answers about installing and using this skill.
Document chunking and Milvus ingestion prepare local files by converting them to Markdown, generating QA-ready frontmatter, and splitting content into chunks for vector database storage. This process produces queryable knowledge assets without requiring external web research.
Batch import local file directories by providing directory paths to the ingestion workflow. The system validates supported formats across the directory, triggers doc-converter for Markdown conversion, and coordinates chunking and Milvus ingestion, returning a summary with per-document results and failure details.
Document ingestion validates input files against supported formats before triggering doc-converter. The doc-converter transforms valid local documents into a common Markdown representation, preparing raw and archived outputs for downstream chunking and Milvus vector database ingestion.
No, the document ingestion workflow processes local files end-to-end without requiring external web research. It handles format validation, Markdown conversion, frontmatter generation, chunking, and Milvus persistence entirely through internal orchestration of doc-converter and knowledge-persistence components.
Failed document ingestion attempts are captured in the returned ingestion summary, which includes per-document results and specific failure details. This enables targeted retry and auditing of individual files without reprocessing the entire batch import directory.
zinohomeConvert PDFs, office files, images, audio, and web content into Markdown.
peterlianpiCreate document embeddings and index them in a vector store for retrieval-augmented generation.
mengsi16Chunk documents, generate QA pairs, and persist to Milvus.
DhumitechImplement end-to-end RAG pipelines for ingestion, embedding, retrieval, and generation.
