Classify pasted material into tags and append source-attributed entries to wiki pages.
npx skills add https://github.com/brandtam/rubber-ducky --skill ingest-writing
Please help me install this Agent Skill. Skill: ingest-writing Source: https://github.com/brandtam/rubber-ducky/tree/main/src/skills/ingest-writing Command: npx skills add https://github.com/brandtam/rubber-ducky --skill ingest-writing
This skill enables you to paste raw material — a draft, an email, or an article — and have it classified, extracted, and routed into the correct context pages (voice / about / vocabulary / preferences). It preserves history by appending new extractions rather than overwriting existing content.
Paste your raw material and ask it to ingest to route extractions into the appropriate context pages.
High-intent search queries and answers about installing and using this skill.
To extract and route pasted material into wiki context pages, paste your raw text and trigger the ingestion workflow. The skill classifies the content into contextual tags and appends new extractions to the correct wiki pages automatically.
Append-only writing for wiki context pages is a non-destructive workflow that adds new extracted information to existing pages without overwriting previous content. It preserves history by appending novel entries with source attribution to relevant sections.
To avoid duplicate entries when ingesting raw text, the workflow reads existing context pages before writing. It compares new extractions against current content and only appends novel information with clear provenance to prevent duplication.
Yes, you can classify pasted drafts into contextual tags like voice, facts, vocabulary, and preferences. The ingestion process identifies these categories within your raw material and routes each extraction to its matching context page.
No, the context page ingestion workflow does not overwrite existing facts. It uses an append-only process that reads current pages to ensure non-destructive updates, preserving all historical entries while adding new source-attributed information.
No, you do not need to format your raw material before routing extractions. You can paste an unformatted draft, email, or article directly, and the skill will classify and extract the relevant content into the appropriate wiki sections.
bahayonghangCompile markdown captures into draft knowledge with summaries and indices.
michaelboedingExtract a structured brand profile from a website into brand_profile.json.
arunodaSearch extracted facts locally using Ollama embeddings and sqlite-vec.
anotherbenParse raw stream-of-consciousness text into categorized vault items with summaries and tags.
