PaperGraph MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_environment_diagnosticsB | Return PaperGraph version and reproducible launch diagnostics. |
| validate_arxiv_inputC | Normalize arXiv ID and URL inputs and return the safe next action. |
| validate_arxiv_requestB | Validate a raw user arXiv request and return the safe next action. |
| open_workspaceC | Open or initialize a persistent multi-paper workspace. |
| workspace_add_local_paperB | Add or replace a local LaTeX project in the active workspace. |
| workspace_add_arxiv_paperB | Add or replace an arXiv LaTeX project in the active workspace. |
| workspace_add_pdf_paperC | Add or replace a born-digital PDF paper in the active workspace. |
| workspace_list_papersB | List all papers stored in the active workspace. |
| workspace_get_paperA | Return metadata and counts for one stored paper. |
| workspace_search_theoremsB | Search theorem titles and bodies across the active workspace. |
| workspace_get_dependenciesC | Return dependencies of a globally identified stored theorem. |
| workspace_get_dependency_diagnosticsC | Explain how workspace dependencies were extracted for one theorem. |
| workspace_get_citationsC | Return incoming or outgoing citation evidence for a stored paper. |
| workspace_list_resultsC | List stored evidence results across the active workspace. |
| workspace_get_resultB | Return one stored evidence result with source spans. |
| workspace_get_result_proofC | Return proof evidence for one stored evidence result. |
| workspace_get_proof_dependenciesC | Return proof dependency evidence for one stored evidence result. |
| workspace_get_external_result_mentionsB | Return external result mentions from a result's proof evidence. |
| workspace_get_evidenceC | Return metadata and source spans for one evidence node or edge. |
| workspace_export_reading_bundleC | Export a paper-level evidence bundle for paper-reading consumers. |
| workspace_export_result_reading_contextB | Export focused evidence context for reading one result's proof. |
| workspace_get_source_sliceC | Return bounded source text around one span, result, or proof. |
| workspace_get_result_reading_pathB | Return deterministic local reading paths for one result. |
| workspace_create_reading_sessionC | Create a persistent reading session in the active workspace. |
| workspace_list_reading_sessionsB | List persistent reading sessions in the active workspace. |
| workspace_get_reading_sessionB | Return one persistent reading session with checkpoints and notes. |
| workspace_record_reading_checkpointC | Create or update a reading checkpoint in the active workspace. |
| workspace_add_reading_noteC | Add a note or question to a reading session. |
| workspace_export_reading_session_summaryC | Export a deterministic recovery summary for a reading session. |
| workspace_create_reading_queueC | Create a persistent reading queue for one stored result. |
| workspace_list_reading_queuesC | List persistent reading queues in the active workspace. |
| workspace_get_reading_queueB | Return one persistent reading queue with ordered items. |
| workspace_apply_reading_queue_to_sessionC | Apply reading queue items as checkpoints in a reading session. |
| workspace_plan_external_imports_for_resultC | Plan external arXiv imports for one result's reading path. |
| workspace_plan_external_imports_for_queueC | Plan external arXiv imports referenced by one reading queue. |
| workspace_plan_external_imports_for_paperC | Plan external arXiv imports visible in one stored paper. |
| workspace_resolve_external_referenceC | Resolve a blocked external reference to a user-confirmed target. |
| workspace_list_external_reference_resolutionsC | List recorded external reference resolutions and selection provenance. |
| workspace_search_external_referenceC | Search scholarly metadata providers for a blocked external reference. |
| workspace_list_external_reference_searchesC | List scholarly reference search runs and candidates. |
| workspace_resolve_external_reference_candidateC | Apply a searched candidate as an explicit reference resolution. |
| workspace_get_paper_mapC | Return an evidence-first first-load map for one stored paper. |
| workspace_export_paper_reading_reportC | Export a deterministic Markdown reading report for one stored paper. |
| workspace_export_cross_paper_reading_planC | Export a deterministic Markdown reading plan for selected papers. |
| workspace_plan_starter_projectC | Plan a Workspace Starter run without writing files. |
| workspace_bootstrap_reading_projectC | Create starter artifacts for a first reading project. |
| load_paperA | Load a local LaTeX paper and build its theorem graph. |
| load_arxiv_paperC | Download an arXiv source project and build its theorem graph. |
| load_arxiv_requestB | Validate a raw arXiv request, then load it only if unambiguous. |
| list_theoremsC | List theorem-like environments in the currently loaded paper. |
| get_theoremB | Return the full text and metadata for one theorem-like node. |
| get_dependenciesC | Return theorem-like nodes referenced by the given theorem. |
| get_dependency_diagnosticsB | Explain how dependencies were extracted for one theorem-like node. |
| where_usedB | Return theorem-like nodes that reference the given theorem. |
| workspace_create_reference_expansionC | Save an approved finite expansion policy. Creation does not search or import. |
| workspace_advance_reference_expansionB | Execute the saved policy: search, download arXiv sources, import unique strong candidates and record resolutions without per-paper prompts, within approved budgets. Resume with the same run ID. |
| workspace_get_reference_expansionB | Read saved graph, counts, decisions and continuation actions without network. |
| workspace_list_reference_expansionsC | List saved expansion summaries, optionally filtered by state. |
| workspace_decide_reference_expansionC | Record exactly one candidate_id, target, existing_paper_id, skip:true or retry:true. Advance separately to execute the approved choice. |
| workspace_update_reference_expansion_policyB | Explicitly revise numeric budgets; cumulative usage is retained. |
| workspace_pause_reference_expansionC | Pause scheduling after the active step. Advance explicitly to resume. |
| workspace_cancel_reference_expansionC | Permanently stop this run, preserving imported papers and history. |
| workspace_export_reference_expansionC | Return saved JSON or Markdown without writing client files or using network. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 63 tools
Many tools have clear roles, but there are several near-duplicate pairs (validate_arxiv_input vs validate_arxiv_request, load_paper/load_arxiv_paper vs workspace_add_*_paper, get_dependencies vs workspace_get_dependencies) that an agent could easily confuse. The external-reference and expansion workflows also use many similar names, making boundaries unclear.
The dominant workspace_verb_noun pattern is readable and consistent, but a significant segment of unprefixed legacy-style tools (load_paper, list_theorems, get_theorem, get_dependencies, where_used, validate_arxiv_*) breaks the convention. It is mixed rather than chaotic, so it still earns a middle score.
With 63 tools, this server is far beyond a typical agent-friendly surface and fits the extreme-count end of the scale. Many tools are workflow-stage variants or duplicates that could be consolidated into fewer, more general operations.
The surface covers the core domain thoroughly: paper ingestion, theorem/dependency extraction, evidence and results, reading sessions/queues, exports, and external reference handling. Minor lifecycle gaps exist (no explicit delete/remove tools for papers, sessions, or queues), but agents can reasonably work around them.