qdrant-rag-pipeline
|
Works with
Claude CodeCursorCodex CLIGitHub CopilotGemini CLI
--- name: qdrant-rag-pipeline description: | license: Apache-2.0 --- # RAG Pipeline with Qdrant A RAG pipeline has three independently tunable stages. Bad answers almost always trace back to exactly one of them: the chunks fed in, the retrieval logic pulling them out, or the context window assembled for the LLM. Fixing the wrong stage wastes time. Diagnose first: are retrieved chunks relevant but answers still wrong (context-assembly problem), are chunks irrelevant (retrieval problem), or are chunks retrieved correctly but lack the content needed to answer (chunking problem)? ## Sub-skills - [chunking](chunking/SKILL.md) — document segmentation strategies, chunk size, overlap, metadata - [retrieval](retrieval/SKILL.md) — query strategies, filters, hybrid search, score thresholds - [context-assembly](context-assembly/SKILL.md) — ranking retrieved chunks, deduplication, prompt packing
More AI & ML skills
writing-shape
mattpocock/skills
Writing, exploit: shape raw material into an article, paragraph by paragraph.
279.3k
writing-fragments
mattpocock/skills
Writing, explore: mine raw fragments, no structure yet.
279.2k
full-output-enforcement
leonxlnx/taste-skill
Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.
275.4k
