Convert PDFs and DOCX files to searchable Markdown with OCR.
npx skills add https://github.com/Guard8-ai/SEAR --skill sear
Please help me install this Agent Skill. Skill: sear Source: https://github.com/Guard8-ai/SEAR/tree/main/claude-skill Command: npx skills add https://github.com/Guard8-ai/SEAR --skill sear
This Skill eliminates the manual effort and high costs associated with searching vast document collections, building RAG applications, and converting complex file formats like PDFs and DOCX. It provides precise, cited answers, saving you time and money by automating knowledge retrieval and document processing.
Convert 'quarterly_report.pdf' to markdown, index it as 'finance_docs', and then search for 'revenue growth projections' within the 'finance_docs' corpus.
High-intent search queries and answers about installing and using this skill.
Semantic search lets you query documents by meaning rather than keywords, returning LLM-synthesized answers with line-level citations. Convert your PDFs and DOCX files to indexed markdown, then perform searches across your corpus to get precise, sourced results without manual document review.
Yes. The Skill converts scanned PDFs to markdown using OCR, preserving structure and metadata. It handles both native and image-based PDFs, and supports language hints to optimize extraction accuracy for multilingual documents.
Index diverse document corpuses using FAISS for semantic indexing, then perform multi-corpus searches that aggregate results across all indexed collections. The Skill synthesizes answers with clear source attribution, enabling retrieval-augmented generation at scale without high API costs.
Content extraction retrieves raw, relevant document sections based on semantic similarity without generation overhead. This is useful for focused analysis, compliance review, or feeding chunks into downstream processing pipelines without incurring LLM costs.
Yes. The Skill runs locally using FAISS and CPU-optimized indexing, supporting offline semantic search without cloud dependencies. GPU acceleration is optional for faster indexing on large corpuses, making it cost-effective for any document scale.
Multi-corpus management lets you organize diverse document types into separate indexes, then search across all of them simultaneously with unified attribution. Each result clearly identifies its source corpus, enabling cross-domain knowledge retrieval for compliance, research synthesis, or codebase analysis.
rysweetExtracts text and tables from PDFs, creates, merges, and handles OCR for scanned documents.
jtucker9Retrieve past CC session details via semantic vector search with LanceDB and SQLite fallback.
OpenCovenPerform vector similarity search using Facebook's FAISS library.
icyzhIndex and search local markdown notes using BM25 and semantic embeddings.
