Documents skills

Browse reusable Agent Skills, each with a clear purpose and practical guidance.

semantic-grep

In-process semantic search over text files or in-memory strings, using Gemini embeddings via the CF AI Gateway. Use when user wants fuzzy/conceptual search where exact-keyword grep would miss — "sessions discussing regulatory constraints", "code about retry logic", "notes mentioning burnout even if the word isn't there". Complements searching-codebases (regex/AST) and extracting-keywords (YAKE). Do NOT use when an exact string/regex match is what's wanted — grep/rg wins on speed and precision there.

129 repo starsObserved in 1 repos

Documents

GitHub

suite-converter

Converts test suites from external eval frameworks into the Margin Eval suite format. Use this skill whenever the user wants to import, convert, translate, or migrate an eval dataset or test suite into Margin Eval format, or when they mention converting tasks from other benchmarking frameworks into Margin's structure.

129 repo starsObserved in 1 repos

Documents

GitHub

transcribing-images

Reads the visual content of slides, pages, and images the way a human would, not just their embedded text. Use when a PPTX or PDF has image slides, screenshots, charts, scanned figures, or flattened-to-image layouts that the built-in pptx/pdf skills read as empty; when asked to transcribe, describe, OCR, or extract what is shown in an image, slide deck, or document page; or when embedded-text extraction returned little or nothing from a visually rich file. Triggers on 'read this deck', 'what's on these slides', 'transcribe', 'OCR', 'extract text from image', 'describe this chart/diagram', .pptx/.pdf/.png/.jpg with visual content.

129 repo starsObserved in 1 repos

Documents

GitHub

whiteboard-video

Generate smooth hand-drawn whiteboard explainer videos directly inside Codex from text scripts, scene plans, SVGs, line art, or local images. Use when the user asks for SpeedPaint-like whiteboard videos, stroke/path drawing, SVG-first rendering, local image-to-line-art extraction, raster skeleton tracing, hand/pen-tip following, color fill, narration/TTS, FFmpeg composition, or script-to-video whiteboard workflows.

129 repo starsObserved in 1 repos

Documents

GitHub

csv-data-analysis

This skill should be used when users need to analyze CSV or Excel files, understand data patterns, generate statistical summaries, or create data visualizations. Trigger keywords include "analyze CSV", "analyze Excel", "data analysis", "CSV analysis", "Excel analysis", "data statistics", "generate charts", "data visualization", "分析CSV", "分析Excel", "数据分析", "CSV分析", "Excel分析", "数据统计", "生成图表", "数据可视化".

128 repo starsObserved in 6 repos

Documents

GitHub

book-ingest

把用户提供的书源文件(epub / mobi / azw / azw3 / prc / md / markdown / 文字型 pdf)提取并按章切分, 在 echo-reading 项目里生成 books/书名/chNN/raw.md(**只灌纯原文**)和 progress.md,并把源文件留档进 books/书名/。当用户给出电子书文件、Markdown、文字型 PDF、下载链接,或说「导入这本书 / 加一本书 / 把这本书切好 / 读这本新书 / 这个文件能不能拆成章」时,务必使用本 skill。PDF 必须先探测;扫描型/图片型 PDF 暂不支持,需停止并告知用户。入库只到原文就位为止。