6 个 GitHub 仓库把 Astra 变成研究智能体
用 Crawl4AI、SearXNG、Docling、Qdrant、Letta、DeepEval 六件套给 GPT-6 Astra 搭检索、记忆和引用校验层。
one model + 6 GitHub repos gets stupid fast enough to turn GPT 6 Astra into a research agent that can actually cite sources 01 Crawl4AI ▸ https://github.com/unclecode/crawl4ai → crawls the web into LLM friendly markdown 02 SearXNG ▸ https://github.com/searxng/searxng → runs private search without locking you to one engine 03 Docling ▸ https://github.com/docling-project/docling → parses PDFs and Office files into clean structured input 04 Qdrant ▸ https://github.com/qdrant/qdrant → stores retrieved chunks for fast similarity search 05 Letta ▸ https://github.com/letta-ai/letta → keeps research memory across sessions so context survives 06 DeepEval ▸ https://github.com/confident-ai/deepeval → checks whether answers stay grounded in the sources Astra should spend tokens on synthesis and judgment Search, parsing, storage, memory and eval stay in dedicated layers That is the bottleneck a model alone never fixes
这条是引用着另一条写的：x.com/promixyz/status/2102424036416348494