bip-extract-refs

npx skills add majiayu000/claude-skill-registry --skill bip-extract-refs --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

# Extract References from LaTeX Repository Analyze a LaTeX repository to extract papers and concepts for the bipartite knowledge graph. ## Usage ``` /bip-extract-refs <path-to-tex-repo> ``` ## What This Skill Does 1. **Reads** `main.bib` and `main.tex` from the specified repository 2. **Extracts** citation keys and their context from the manuscript 3. **Identifies** key concepts and potential relationships 4. **Filters out** software/tool citations (e.g., packages, libraries) 5. **Returns** a structured summary with: - Papers found (using BibTeX citation keys as bip IDs) - Suggested concepts to create - Potential edges to add ## Workflow Given a path like `../dasm-tex-1`, perform this analysis: ### Step 1: Read the BibTeX file ```bash cat <repo>/main.bib ``` Extract citation keys (the part after `@article{` or `@inproceedings{`, etc.). ### Step 2: Read the manuscript ```bash cat <repo>/main.tex ``` Look for: - `\cite{...}` commands and their surrounding context - Section structure to understand paper focus - Key methodological terms ### Step 3: Identify Concepts Look for recurring themes that could become concept nodes: - Methods mentioned repeatedly - Domain-specific terms - Mathe

cat <repo>/main.bib
cat <repo>/main.tex
Check which papers exist in bip
bip get Smith2026-ab
Create suggested concepts
bip concept add x-model --name "X Model" --description "..."
Add edges
bip edge add -s Smith2026-ab -t x-model -r introduces -m "Foundational paper"