Loading...
Loading...
Found 82 Skills
Converts an input document (.txt, .md, .pdf, .docx) into a structured RSVP token stream with ORP alignment and configurable WPM. Use when a user wants to speed-read a document, prepare a reading session, or generate a token stream for a speed-reading UI.
Build or rebuild the ADR index and dependency graph in AgentDB
Extract text from PDFs/scans (pymupdf, marker-pdf).
Extract structured fields, original text locations, missing items, and supportable judgments from papers and supplementary materials. Use when the user asks for "extract this paper", "batch extract paper information", "create paper cards", or requests the rw-paper-extractor workflow. Runs without a private local workspace or preset research-lab; use user-provided material and bundled public-source methods.
Download PDFs (when available) and extract plain text to support full-text evidence, writing `papers/fulltext_index.jsonl` and `papers/fulltext/*.txt`. **Trigger**: PDF download, fulltext, extract text, papers/pdfs, 全文抽取, 下载PDF. **Use when**: `queries.md` 设置 `evidence_mode: fulltext`（或你明确需要全文证据）并希望为 paper notes/claims 提供更强 evidence。 **Skip if**: `evidence_mode: abstract`（默认）；或你不希望进行下载/抽取（成本/权限/时间）。 **Network**: fulltext 下载通常需要网络（除非你手工提供 PDF 缓存在 `papers/pdfs/`）。 **Guardrail**: 缓存下载到 `papers/pdfs/`；默认不覆盖已有抽取文本（除非显式要求重抽）。
Automate PDF API IO tasks via Rube MCP (Composio). Always search tools first for current schemas.
Deconstruct academic papers into core contributions, critical assumptions, and napkin-worthy insights. Use when analyzing research papers.
OCR Web Service integration. Manage Documents. Use when the user wants to interact with OCR Web Service data.
Use markdown.new to fetch web content, convert files, and crawl sites as clean Markdown instead of raw HTML. Trigger this skill whenever you need to read a webpage, fetch URL content, convert a document (PDF, DOCX, XLSX, images, etc.), or crawl a website. This includes any task involving summarizing web pages, extracting content from URLs, ingesting documents, researching topics online, or gathering information from websites. If you're about to fetch HTML from a URL, use markdown.new instead — it returns clean Markdown with 80% fewer tokens.
Processes a new source from raw/ into the wiki. Activates when the user asks to ingest, process, add, or incorporate a source into the knowledge base.
Summarize text content, web pages, documents, and long-form articles into concise, structured summaries.
Extract methodologies from documents or examples to create executable skills