wiki-research

Researches topics via multi-round web search and files structured findings into an Obsidian wiki.

Updated Nov 9, 2023
One-click install
npx skills add https://github.com/oresttokovenko/dot-files --skill wiki-research-oresttokovenko
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wiki-research
Source: https://github.com/oresttokovenko/dot-files/tree/main/chezmoi/dot_agents/skills/wiki-research
Command: npx skills add https://github.com/oresttokovenko/dot-files --skill wiki-research-oresttokovenko

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Manually researching a topic across the web and organizing findings into a personal knowledge base is slow and inconsistent. This Skill automates multi-round web research and files synthesized, cross-linked results directly into an Obsidian vault. ## Core Features & Use Cases - Multi-Round Autonomous Research: Runs up to three rounds of web searches — broad survey, gap fill, and contradiction resolution — extracting claims, concepts, entities, and disagreements from sources. - Structured Wiki Filing: Writes source, concept, entity, and synthesis pages with provenance and confidence frontmatter, then cross-links everything and updates index.md, log.md, hot.md, and .manifest.json. - Pluggable Retrieval Backends: Optionally loads CLI backends (yt-dlp transcripts, defuddle, Perplexity) from a vault registry with free-first ordering and graceful fallback to WebSearch/WebFetch. - Use Case: Ask for a deep dive on "vector databases" and receive a master synthesis page plus linked source, concept, and entity pages filed permanently in your Obsidian wiki. ## Quick Start Ask the assistant to run /wiki-research on a topic such as "do a deep dive on vector databases and file the results into my wiki".

Frequently Asked Questions about wiki-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run autonomous web research into an Obsidian vault?

Invoke the skill with a topic, for example "/wiki-research vector databases". It resolves your vault config, runs up to three rounds of web searches, and writes synthesis, source, concept, and entity pages directly into the vault.

What output does a multi-round research workflow produce?

It produces a master synthesis page at synthesis/Research: [Topic].md plus individual source, concept, and entity pages. All pages carry provenance and confidence frontmatter and are cross-linked, with index.md, log.md, hot.md, and .manifest.json updated.

Can I use CLI tools like yt-dlp or Perplexity as research backends?

Yes, define them in references/research-backends.md in your vault as a YAML backend list. The skill checks binary availability and API keys, prefers free backends first, and falls back to WebSearch/WebFetch on any failure.

Does the research skill avoid duplicating existing wiki pages?

Yes, it reads index.md before starting to see what the wiki already covers and checks for existing pages when filing. Findings are merged into existing pages rather than creating duplicates.

What happens when sources contradict each other during research?

Contradictions are tracked across rounds, and a final targeted pass attempts authoritative resolution. If resolution is impossible, the disagreement is explicitly flagged in the synthesis page's Contradictions & Open Questions section.