synthesize obsidian research

The problem

Research vaults in Obsidian often become "write-only" graveyards of fragmented highlights, ambiguous paraphrases, and untraceable claims. Developers and researchers struggle to distinguish their own ideas from source material, leading to weak arguments and evidence gaps.

What it does

• Performs a read-only health diagnostic to identify fragmentation, link quality, and citation coverage.
• Maps existing notes into a structured research model including source notes, evidence items, and claims.
• Constructs a claim ledger that links propositions to specific supporting or contradicting evidence with exact locators.
• Generates synthesis reports that highlight source disagreement and identify missing data points.
• Prepares reversible, incremental vault updates to improve organization without destroying original wording.

Frameworks & tools

Designed for Obsidian vaults using Markdown. Utilizes citation keys, BibTeX/Zotero conventions, and YAML frontmatter. Includes a Python-based vault scanner for local execution.

Why this beats prompting it yourself

Generic LLM prompts often hallucinate connections or "average away" source disagreements. This skill enforces strict evidence traceability, preventing the AI from turning semantic similarity into false evidential support while preserving your specific citation keys and folder structures.

Use cases

• Auditing a research vault for evidence gaps before starting a thesis or whitepaper.
• Consolidating messy Zotero imports into a structured literature review.
• Building a defensible claim-evidence matrix for technical reports or books.
• Mapping contradictions between multiple sources in a specialized domain.