semantic-search
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CORPUS_ROOT | No | Path to the text-layer corpus root directory. Can also be passed as a command-line argument to build_index.py and server.py. |
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 |
|---|---|
| semantic_searchA | Semantic recall: retrieve text-layer passages by meaning, catching what keyword grep misses (paraphrase and cross-lingual matches). Returns top-k fragments, each with file (relative path), page_start/page_end (for going back to the source PDF), score (cosine; higher = closer), and text (the fragment itself). ā Recall layer only: verbatim quotes, page numbers, and emphasis must be verified against the source PDF. Never cite these fragments directly. |
| semantic_search_infoA | Report current index state: backend, model, chunk count, file count, build time. |
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 2 tools
The two tools have completely distinct purposes: one performs semantic search, the other reports index state. No ambiguity.
Both tools follow a consistent snake_case pattern (semantic_search, semantic_search_info).
Only two tools for a semantic search server feels thin; a typical search server would include indexing or configuration tools. The scope is borderline minimal.
Crucial operations like adding or updating documents are missing. The server is read-only and lacks full lifecycle coverage for semantic search.