关于
The `learn-datalake` skill is a continuous orchestrator that monitors a directory, processes new PDFs through quality review loops, and ingests other file types into graph memory. It automatically extracts and links framework controls (like NIST or ATT&CK) from PDF content, enabling semantic search and multi-hop traversal. Use this skill to automatically build a queryable knowledge graph from a watched folder of documents and assets.
快速安装
Claude Code
推荐npx skills add grahama1970/agent-skills -a claude-code/plugin add https://github.com/grahama1970/agent-skillsgit clone https://github.com/grahama1970/agent-skills.git ~/.claude/skills/learn-datalake在 Claude Code 中复制并粘贴此命令以安装该技能
GitHub 仓库
Frequently asked questions
What is the learn-datalake skill?
learn-datalake is a Claude Skill by grahama1970. Skills package instructions and resources that Claude loads on demand, so Claude can perform learn-datalake-related tasks without extra prompting.
How do I install learn-datalake?
Use the install commands on this page: add learn-datalake to Claude Code as a plugin, or clone its repository into your skills directory, then restart Claude so it picks up the skill.
What category does learn-datalake belong to?
learn-datalake is in the Documents category, tagged pdf and data.
Is learn-datalake free to use?
Yes. learn-datalake is listed on AIMCP and free to install. It runs inside Claude, so no separate service account is required to use the skill itself.
相关推荐技能
这个Skill为开发者提供了语义化版本规范和变更日志格式标准。它能在准备软件发布时快速指导版本号更新和变更日志撰写，包含版本号递增规则、预发布标识符等关键信息。适用于需要遵循规范发布流程的开发场景。
这个Skill帮助开发者遵循Conventional Commits规范格式化Git提交信息。它提供了标准格式模板和常用提交类型的中英文对照表（如feat/新增、fix/修正等），适用于编写提交、执行git commit或审查提交历史的场景。通过确保提交信息的规范性和一致性，它能提升团队协作效率和版本历史可读性。
HuggingFace Tokenizers 提供了基于 Rust 的高性能分词工具，支持 BPE、WordPiece 和 Unigram 算法，能在一分钟内处理 1GB 文本。它适用于需要快速分词或训练自定义词汇表的场景，并能无缝集成到 transformers 库中。开发者可以借助它进行对齐跟踪、填充截断等操作，满足从研究到生产的全流程需求。
nano-pdf 让开发者能用自然语言指令直接编辑PDF文件，无需手动操作复杂工具。它通过命令行快速修改指定页面内容，如修正拼写错误或更新标题，适合处理日常文档微调。使用前请注意核对页码和输出结果，确保修改准确无误。