abyss-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CHUNK_SIZE | No | Chunk size for chunker (planned). | 512 |
| CHUNK_TOP_K | No | Top K for ranker (planned). | 3 |
| SEARXNG_URL | No | The URL of the SearXNG instance to use for search. | http://127.0.0.1:4243 |
| SEARCH_TOP_K | No | Number of top search results to return. | 5 |
| CHUNK_OVERLAP | No | Chunk overlap for chunker (planned). | 64 |
| CRAWL_TIMEOUT | No | Timeout for crawling in seconds (planned). | 6.0 |
| SEARCH_TIMEOUT | No | Timeout for search requests in seconds. | 8.0 |
| CRAWL_MAX_BYTES | No | Maximum bytes to crawl (planned). | 1500000 |
| CRAWL_CONCURRENCY | No | Concurrency for crawling (planned). | 3 |
| SEARCH_CATEGORIES | No | Search categories to use. | general |
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 |
|---|---|
| searchA | Search the web for current information. Returns the top results as URL and snippet. |
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 1 tool
There is only one tool, so there is no possibility of an agent confusing it with another action. Its purpose ('search the web') is unambiguous.
The single name 'search' is a clean, conventional verb choice, but with only one tool there is no pattern to demonstrate consistency against.
A one-tool server is inherently thin; even for a narrow web-search purpose, a single tool leaves no room for query variants, site scoping, or content retrieval.
Search returning only URLs and snippets covers the stated core purpose, but there is no way to fetch or read page contents, which is a common dead end for an agent needing current information.