Struct ApiClient 

pub struct ApiClient { /* private fields */ }

Implementations

impl ApiClient

pub fn new(config: Config) -> Self

pub async fn execute( &self, endpoint: &EndpointRecord, args: &Value, auth_manager: &mut AuthManager, ) -> Result<Value>

Executes endpoint against the configured SQL Server instance. auth_manager resolves to a tiberius::AuthMethod (see auth::auth_manager::AuthManager::resolve_tds_auth) from the server’s own configured credentials – there is no per-request credential override; SQL Server auth is always this server’s own configured identity, not something a caller supplies per call.

A top-level database string in args (a sibling of body, e.g. {"database": "reporting", "body": {...}}) lets a caller target a specific database for a sandbox-tagged operation, overriding the connection’s own current database (see build_statement’s doc comment for why “sandbox” needs this at all). It’s not part of the generated/documented input schema – mcpify’s schema wrapper doesn’t set additionalProperties: false, so an extra key here passes validate_input unnoticed – and it’s a no-op for master/msdb operations, which are always sent against their real, literal database regardless of this argument.