mcp-document-intelligence
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| scan_directoryA | Scan a directory for documents. Returns filename, path, size, and optionally a text preview (OCR-enabled). Use listOnly=true for fast directory listing without text extraction. Supports PDF, DOCX, DOC, Pages, images, TXT, RTF, ODT. |
| analyze_documentA | Deep-analyze a single document: extract text (with OCR fallback), detect dates, references, keywords, and suggest an intelligent filename. Optionally uses Perplexity AI for classification. |
| analyze_folderA | Batch-analyze all documents in a folder. Returns metadata, filename suggestions, duplicate detection, and performance metrics. |
| rename_documentB | Safely rename a document with collision protection and undo tracking. |
| move_documentB | Move a document to a different directory with collision protection and undo tracking. Creates target directory if needed. |
| smart_renameB | AI-powered smart rename: analyzes document content and renames it intelligently based on detected date, company, document type, and reference numbers. Uses Perplexity AI when configured, falls back to pattern matching. |
| find_folderB | Fuzzy-search for a folder by name. Supports typo tolerance via Levenshtein distance and returns ranked matches. |
| batch_organizeB | Batch rename and move files according to a list of operations. Supports move/copy mode, automatic backup for undo, and detailed reporting. |
| auto_organizeA | Autonomously analyze and organize all documents in a source folder — combines analysis + rename + move in one step. |
| process_downloadsB | Scan the downloads folder, analyze documents, and sort them into an archive. Supports auto-move or preview mode. |
| undo_operationA | Undo the last batch of rename/move operations. Restores original filenames and locations. |
| export_metadataC | Export document metadata to JSON or CSV format. |
| suggest_structureA | Suggest an intelligent folder structure based on analyzed documents — groups by year, category, and company. |
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 13 tools
Most tools have clear distinct roles, but the organize family — auto_organize, process_downloads, batch_organize, and analyze_folder — overlaps heavily, differing mostly in scope (downloads vs arbitrary folder vs pre-supplied list). Similarly, rename_document vs smart_rename is only distinguished by the AI hint in the description, which an agent could overlook.
All names are snake_case and use action verbs (scan_, analyze_, rename_, move_, find_, undo_, export_), which is predictable. Minor deviations: smart_rename, auto_organize, and batch_organize prefix the verb with a modifier, breaking the pure verb_noun pattern.
13 tools is squarely in the reasonable range for a document intelligence server. There is mild redundancy among the three organize/pipeline tools, but each still earns a place for a distinct workflow.
The document lifecycle is well covered: scan, analyze (single and batch), rename, move, organize, undo, find, and export metadata. Gaps remain for delete/discard operations and a standalone copy or preview/get-document tool, so agents must work around a few edges.