# AI use cases for Papergraph MCP (lotchuazzz-crypto/papergraph-mcp) Repo: [lotchuazzz-crypto/papergraph-mcp](https://github.com/lotchuazzz-crypto/papergraph-mcp) | Use cases: 4 | Page: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/ PaperGraph MCP is a Python tool that turns arXiv papers, local LaTeX projects, and born-digital PDFs into a local, evidence-grounded workspace of math results, proofs, and dependencies for AI agents and readers. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Map a math paper's main results before reading: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/map-math-paper-results/ - Trace a theorem's dependencies before reading its proof: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/trace-theorem-dependencies/ - Find public full text for a DOI paper: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/find-public-full-text-for-doi/ - Export a Reading Report and cross-paper reading plan: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/export-reading-report-and-plan/ ## How to map a math paper's main results before reading with Papergraph MCP Repo: [lotchuazzz-crypto/papergraph-mcp](https://github.com/lotchuazzz-crypto/papergraph-mcp) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Get a Paper Map of structure, main-result candidates, and external reading risks before you pick a theorem. **Who it is for:** A graduate student or researcher who has a new math paper and wants to know which theorem to read first. **The problem:** Dense papers hide which results matter and what outside material they rely on, so readers often start in the wrong place. **The outcome:** A Paper Map listing the paper structure, main-result candidates, and external reading risks, so the reader can choose a theorem with source evidence. ### Steps 1. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp doctor` to confirm the install works. 2. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp plan-starter-project` to preview and validate your input files before anything is written. (inferred, check the README) 3. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp bootstrap-reading-project` after reviewing the preview to create the workspace. (inferred, check the README) 4. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp get-paper-map --workspace WORKSPACE --paper-id PAPER_ID` and review the main-result candidates. ### Prompt to paste into your agent ```text Use the papergraph-mcp repo to build a Paper Map for the paper at in the workspace at . Return the main-result candidates and external reading risks as a ranked list. Verify that each candidate cites a source slice from the paper. ``` ### Caveat PaperGraph does not verify proofs, so the Paper Map ranks candidates for reading and does not confirm which result is correct. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/map-math-paper-results/ ## How to trace a theorem's dependencies before reading its proof with Papergraph MCP Repo: [lotchuazzz-crypto/papergraph-mcp](https://github.com/lotchuazzz-crypto/papergraph-mcp) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README See direct and recursive dependencies, shared prerequisites, and unknowns for one chosen result. **Who it is for:** A researcher preparing to read or teach a specific theorem who wants to know what they need to understand first. **The problem:** Proofs cite earlier lemmas and outside results, and tracing those references by hand is slow and error-prone. **The outcome:** A compact dependency reading for the chosen result showing direct and recursive dependencies, shared prerequisites, cycles, and unknowns, with full evidence still available. ### Steps 1. Open the Paper Map for the paper and choose a result ID from the main-result candidates. 2. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp get-dependency-reading WORKSPACE PAPER_ID --target-result-id RESULT_ID --detail compact`. 3. Check the listed unknowns and cycles, and open the source slices for any dependency you plan to rely on. 4. Record your open questions and checkpoints in the workspace before moving to the next result. ### Prompt to paste into your agent ```text Using the papergraph-mcp repo, run get-dependency-reading on the paper in workspace for result . Report the direct dependencies, shared prerequisites, and unknowns. Verify that every listed dependency links to a source slice and that unknowns are not presented as established. ``` ### Caveat Empty dependencies do not prove a result is independent, and PaperGraph does not infer hidden prerequisites. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/trace-theorem-dependencies/ ## How to find public full text for a DOI paper with Papergraph MCP Repo: [lotchuazzz-crypto/papergraph-mcp](https://github.com/lotchuazzz-crypto/papergraph-mcp) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Discover public PDF candidates for a DOI and import the root paper when one eligible copy exists. **Who it is for:** A reader who has a DOI from a citation list and needs a readable public copy of the paper. **The problem:** A DOI resolves to a landing page, not a readable paper, and finding a legal public full-text copy takes manual searching. **The outcome:** A list of public full-text candidates for the DOI, and an imported root paper when exactly one eligible public PDF exists. ### Steps 1. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp discover-doi DOI` to list candidates without downloading anything. (inferred, check the README) 2. Review the candidates and choose one if several are returned, since multiple candidates require your selection. 3. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp add-doi-paper WORKSPACE DOI` to import the root paper when exactly one eligible public PDF exists. (inferred, check the README) 4. Run `uvx --from papergraph-mcp==1.2.3 papergraph-mcp get-paper-map --workspace WORKSPACE --paper-id PAPER_ID` on the imported paper. ### Prompt to paste into your agent ```text Use the papergraph-mcp repo to discover public full-text candidates for DOI . Return the candidates with their sources and do not import anything until the user chooses one. Verify that no file was downloaded during discovery. ``` ### Caveat Metadata is not full text, and PaperGraph does not bypass paywalls, so some DOI papers will have no importable copy. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/find-public-full-text-for-doi/ ## How to export a Reading Report and cross-paper reading plan with Papergraph MCP Repo: [lotchuazzz-crypto/papergraph-mcp](https://github.com/lotchuazzz-crypto/papergraph-mcp) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Save checkpoints and notes, then export Markdown reports that summarize your reading across papers. **Who it is for:** A researcher reading several related papers who needs a written record of progress and a suggested order. **The problem:** Reading progress, notes, and open questions get scattered across papers, making it hard to resume or hand off work. **The outcome:** A Reading Report for each paper and a Cross-Paper Reading Plan in Markdown, backed by the local workspace state. ### Steps 1. Save checkpoints, notes, and open questions in the workspace as you read each result. 2. Follow the export commands in the first-workspace walkthrough to produce a Reading Report. 3. Export a Cross-Paper Reading Plan once two or more papers are in the same workspace. 4. Keep the workspace database outside the repository and back it up before any schema upgrade. ### Prompt to paste into your agent ```text Using the papergraph-mcp repo and workspace , export a Reading Report and a Cross-Paper Reading Plan. Return both as Markdown files and check that each open question in the report appears in the workspace state. ``` ### Caveat Papers and reading state stay local, and older PaperGraph versions cannot open schema 9 workspaces, so back up before upgrading. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/lotchuazzz-crypto-papergraph-mcp/export-reading-report-and-plan/