
Learn: Multi-Source Research Workflow
Turns scattered sources into a structured, publish-ready reference with a six-phase research process.
- Repository
- tw93/Waza
- Stars
- 7,493
- Licence
- MIT
- Files
- 1 file
What it does
When installed, the agent follows a six-phase process to research an unfamiliar topic: it collects primary sources, takes notes, builds an outline, drafts sections, edits, and reviews the result. It checks for contradictions, flags uncertainty, and can produce research notes, an outline, a draft article, or a single canonical reference, depending on the mode chosen. It is designed to support your thinking rather than replace it.
Great for
- Someone learning a new domain who wants a reliable mental model and reference
- Writers turning a pile of articles, papers, or notes into one structured piece
- Anyone who wants one comprehensive reference that makes further searching unnecessary
Skip it if
- Quick factual lookups or reading a single file, which the skill itself says to avoid because the process is heavy
- Situations needing fast answers, since collecting and cross-checking many sources takes time
- Sensitive or paywalled material handled without the companion /read and /write tools, where coverage is reduced and the output may be incomplete
Try asking
- “Help me understand the current state of retrieval-augmented generation; compile sources into a reference”
- “Compile these five research papers and notes into a publish-ready article”
- “Do a deep dive on zero-knowledge proofs and make a one-article reference anyone can read”
- “Give me a quick reference on Kubernetes networking with no article planned”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
learn.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip learn.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip learn.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- State your intended output (notes, outline, draft, or canonical article) up front so the agent picks the right mode without asking
- Supply your own source files or URLs where you can, since the agent prioritises primary sources and you control what is included
- Review the final draft yourself; the skill explicitly expects a human to read the article before publishing
What's inside
- SKILL.md
Redistributed unmodified from tw93/Waza under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.