folder-index
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| root | Yes | Path of the folder to index, passed as `--root <folder>` (e.g. `folder-index --root C:\Users\me\Downloads serve`). |
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
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| search_filesA | Full-text search over file names, paths and extracted contents (PDF, Word, Excel, PowerPoint, text, code, zip listings, OCR'd scans and images). Every word must match; words are prefix-matched. FTS5 syntax ("exact phrase", OR, NOT) also works. ext: comma list like "pdf,docx". folder: substring of the path, or "/..." to search one indexed folder. kind: document | text | archive | image | media | binary | other. modified_after: YYYY-MM-DD. |
| find_filesC | List files by metadata. name: substring or glob (e.g. "invoice2025*"). sort: modified | oldest | size | name. Dates are YYYY-MM-DD. Also reports total count/size. |
| read_file_textB | Return the already-extracted text of an indexed file (path as shown in results, or an absolute path). Page through long files with offset. |
| recent_filesC | Most recently modified/downloaded files. |
| find_duplicatesB | Groups of byte-identical documents (also across indexed folders), largest waste first. |
| index_statusC | Indexed folders, file counts, last update, and update/OCR progress. |
| ocr_speedA | Control the OCR pass over scanned PDFs and images. mode: "turbo" = parallel workers on all CPU cores but one (capped by free RAM), normal priority, fastest; "background" = one low-priority worker (default after updates); "pause" = stop OCR until resumed. workers: override the worker count (0 = auto). A running pass hands its remaining queue over without redoing finished files. |
| refresh_indexA | Start an incremental re-index of every folder in the background (only new/changed files are read). Scanned PDFs and images found along the way are OCR'd afterwards. |
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 8 tools
Most tools have distinct purposes, with search, listing, and reading being clearly separated. However, 'find_files' and 'search_files' share strong metadata filtering capabilities, which could lead to confusion for the agent.
The naming convention is functional but inconsistent. While several tools use clear verb_noun structures (e.g., find_files), others are functional nouns or compound verbs (e.g., index_status, ocr_speed). This mixing makes it less predictable for the agent.
With 9 tools, the server strikes an excellent balance. It covers the entire lifecycle of file indexing, search, and management without becoming overly complex or introducing trivial functions.
The tool surface provides solid coverage for the core functions of a file index, including metadata discovery, full-text search, OCR control, and duplicate detection. Minor gaps exist, such as explicit deletion functions, but the overall workflow is strongly supported.