Literature Evidence MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| XDG_CONFIG_HOME | No | Base directory for user-specific configuration files. If LITERATURE_MCP_CONFIG is not set, the configuration file is expected at $XDG_CONFIG_HOME/literature-mcp/config.json. Defaults to ~/.config. | |
| LITERATURE_MCP_CONFIG | No | Path to the literature-mcp configuration JSON file. Overrides the default configuration location. Can also be specified with the global --config command-line argument. |
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 |
|---|---|
| library_statusA | Check library health, PDF/full-text coverage and enabled capabilities. Does not load writing/research context or experiment records. |
| list_filesA | Browse the local literature folder. folder is relative; results contain file: IDs for fetch. Hidden files and paths outside the configured folder are excluded. |
| searchA | Search live Zotero metadata, local file names and cached PDF/text content. Returns IDs for fetch, snippets and indexing coverage. Uncached PDFs are not full-text searched; query terms are matched lexically on the same page. |
| fetchA | Read a source excerpt by file:, zotero:, project: or feedback: ID. Defaults to one PDF page and 4000 text characters; max_chars is 100..12000. Select at most 50 pages. Continue only as needed, retaining page range and expected_sha256. Use get_writing_context for project IDs. Parent Zotero items auto-select only when one readable PDF exists. |
| read_page_imageA | Inspect an original PDF page as an image, especially for plots, tables, equations or scanned pages. page is 1-based. Requires locally installed Poppler. |
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 5 tools
Each tool maps to a distinct function: health/status, file enumeration, search across metadata and full text, targeted excerpt retrieval, and visual page inspection. The only close pair, fetch and read_page_image, is clearly differentiated as text versus image access.
Names are consistently lowercase snake_case but mix conventions: list_files and read_page_image follow verb_noun, while library_status is noun_noun and search/fetch are bare verbs. The set remains readable but does not follow one predictable naming pattern.
Five tools is appropriately sized for a literature retrieval server; each covers a distinct need without redundancy. The count is well within the ideal 3-15 range and feels neither sparse nor bloated.
The surface covers the core evidence workflow: locate sources via search/list, retrieve text via fetch, and inspect visual content via read_page_image. Minor gaps exist, such as no dedicated full-metadata display or comparison tool, but these are workable.