We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/oculairmedia/Letta-MCP-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
mod connection;
mod error;
mod go_away;
mod peer;
mod ping_pong;
mod settings;
mod streams;
pub(crate) use self::connection::{Config, Connection};
pub use self::error::{Error, Initiator};
pub(crate) use self::peer::{Dyn as DynPeer, Peer};
pub(crate) use self::ping_pong::UserPings;
pub(crate) use self::streams::{DynStreams, OpaqueStreamRef, StreamRef, Streams};
pub(crate) use self::streams::{Open, PollReset, Prioritized};
use crate::codec::Codec;
use self::go_away::GoAway;
use self::ping_pong::PingPong;
use self::settings::Settings;
use crate::frame::{self, Frame};
use bytes::Buf;
use tokio::io::AsyncWrite;
pub type PingPayload = [u8; 8];
pub type WindowSize = u32;
// Constants
pub const MAX_WINDOW_SIZE: WindowSize = (1 << 31) - 1; // i32::MAX as u32
pub const DEFAULT_REMOTE_RESET_STREAM_MAX: usize = 20;
pub const DEFAULT_LOCAL_RESET_COUNT_MAX: usize = 1024;
// RFC 9113 suggests allowing at minimum 100 streams, it seems reasonable to
// by default allow a portion of that to be remembered as reset for some time.
pub const DEFAULT_RESET_STREAM_MAX: usize = 50;
// RFC 9113#5.4.2 suggests ~1 RTT. We don't track that closely, but use a
// reasonable guess of the average here.
pub const DEFAULT_RESET_STREAM_SECS: u64 = 1;
pub const DEFAULT_MAX_SEND_BUFFER_SIZE: usize = 1024 * 400;