/**
* Default configuration values for the scraping pipeline and server
*/
/** Maximum number of pages to scrape in a single job */
export const DEFAULT_MAX_PAGES = 1000;
/** Maximum navigation depth when crawling links */
export const DEFAULT_MAX_DEPTH = 3;
/** Maximum number of concurrent page requests */
export const DEFAULT_MAX_CONCURRENCY = 3;
/** Default protocol for the MCP server */
export const DEFAULT_PROTOCOL = "auto";
/** Default port for the HTTP protocol */
export const DEFAULT_HTTP_PORT = 6280;
/** Default port for the Web UI */
export const DEFAULT_WEB_PORT = 6281;
/** Default host for server binding */
export const DEFAULT_HOST = "127.0.0.1";
/**
* Default timeout in milliseconds for page operations (e.g., Playwright waitForSelector).
*/
export const DEFAULT_PAGE_TIMEOUT = 5000;
/**
* Maximum number of retries for HTTP fetcher requests.
*/
export const FETCHER_MAX_RETRIES = 6;
/**
* Base delay in milliseconds for HTTP fetcher retry backoff.
*/
export const FETCHER_BASE_DELAY = 1000;
/**
* Default chunk size settings for splitters
*/
export const SPLITTER_MIN_CHUNK_SIZE = 500;
export const SPLITTER_PREFERRED_CHUNK_SIZE = 1500;
export const SPLITTER_MAX_CHUNK_SIZE = 5000;
/**
* Maximum number of documents to process in a single batch for embeddings.
*/
export const EMBEDDING_BATCH_SIZE = 100;
/**
* Maximum total character size for a single embedding batch request.
* This prevents "413 Request entity too large" errors from embedding APIs.
* Default is 50000 (~50KB).
*/
export const EMBEDDING_BATCH_CHARS = 50000;
/**
* Maximum number of retries for database migrations if busy.
*/
export const MIGRATION_MAX_RETRIES = 5;
/**
* Delay in milliseconds between migration retry attempts.
*/
export const MIGRATION_RETRY_DELAY_MS = 300;
/**
* Factor to overfetch vector and FTS candidates before applying Reciprocal Rank Fusion.
* A factor of 2 means we fetch 2x the requested limit from each source before ranking.
*/
export const SEARCH_OVERFETCH_FACTOR = 2;
/**
* Weight applied to vector search scores in hybrid search ranking.
*/
export const SEARCH_WEIGHT_VEC = 1.0;
/**
* Weight applied to full-text search scores in hybrid search ranking.
*/
export const SEARCH_WEIGHT_FTS = 1.0;
/**
* Multiplier to cast a wider net in vector search before final ranking.
* Used to increase the number of vector search candidates retrieved.
*/
export const VECTOR_SEARCH_MULTIPLIER = 10;
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/arabold/docs-mcp-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server