diff --git a/Cargo.lock b/Cargo.lock
index a991cc3..1d9cb73 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -151,6 +151,10 @@ dependencies = [
  "syn",
 ]
 
+[[package]]
+name = "ruin_reactivity"
+version = "0.1.0"
+
 [[package]]
 name = "smallvec"
 version = "1.15.1"
diff --git a/lib/runtime/src/channel/mod.rs b/lib/runtime/src/channel/mod.rs
index b58519e..99a62c8 100644
--- a/lib/runtime/src/channel/mod.rs
+++ b/lib/runtime/src/channel/mod.rs
@@ -1,4 +1,14 @@
 //! Async channels for inter-thread communication.
+//!
+//! The channel types in this module are userspace synchronization primitives. They do not carry
+//! messages through the kernel; instead, channel state lives in shared Rust data structures and
+//! the runtime uses `io_uring` `MSG_RING` notifications only to wake the owning runtime thread
+//! when an async waiter becomes ready.
+//!
+//! The initial surface includes:
+//!
+//! - [`oneshot`] for single-value handoff
+//! - [`mpsc`] for bounded and unbounded multi-producer/single-consumer queues
 
 pub mod mpsc;
 pub mod oneshot;
diff --git a/lib/runtime/src/channel/mpsc.rs b/lib/runtime/src/channel/mpsc.rs
index db4634c..f47273d 100644
--- a/lib/runtime/src/channel/mpsc.rs
+++ b/lib/runtime/src/channel/mpsc.rs
@@ -1,3 +1,5 @@
+//! Multi-producer, single-consumer channels.
+
 use std::collections::VecDeque;
 use std::future::poll_fn;
 use std::pin::Pin;
@@ -7,6 +9,13 @@ use std::task::{Context, Poll};
 use crate::op::completion::{CompletionFuture, CompletionHandle};
 use crate::sys::linux::channel::runtime_waiter;
 
+/// Creates a bounded channel with room for at most `capacity` queued messages.
+///
+/// Bounded senders provide both [`Sender::try_send`] and async [`Sender::send`] backpressure.
+///
+/// # Panics
+///
+/// Panics if `capacity == 0`.
 pub fn channel<T: Send + 'static>(capacity: usize) -> (Sender<T>, Receiver<T>) {
     assert!(capacity > 0, "bounded channels require capacity > 0");
     let shared = Arc::new(Mutex::new(State::new(Some(capacity))));
@@ -18,6 +27,9 @@ pub fn channel<T: Send + 'static>(capacity: usize) -> (Sender<T>, Receiver<T>) {
     )
 }
 
+/// Creates an unbounded channel.
+///
+/// Unbounded senders never wait for capacity, but the single receiver is still asynchronous.
 pub fn unbounded_channel<T: Send + 'static>() -> (UnboundedSender<T>, Receiver<T>) {
     let shared = Arc::new(Mutex::new(State::new(None)));
     (
@@ -28,14 +40,17 @@ pub fn unbounded_channel<T: Send + 'static>() -> (UnboundedSender<T>, Receiver<T
     )
 }
 
+/// Bounded multi-producer sender.
 pub struct Sender<T: Send + 'static> {
     shared: Arc<Mutex<State<T>>>,
 }
 
+/// Unbounded multi-producer sender.
 pub struct UnboundedSender<T: Send + 'static> {
     shared: Arc<Mutex<State<T>>>,
 }
 
+/// Single consumer for both bounded and unbounded MPSC channels.
 pub struct Receiver<T: Send + 'static> {
     shared: Arc<Mutex<State<T>>>,
 }
@@ -57,17 +72,24 @@ struct SendWaiter<T: Send + 'static> {
 }
 
 #[derive(Debug, Eq, PartialEq)]
+/// Error returned when sending fails because the receiver has been closed or dropped.
 pub struct SendError<T>(pub T);
 
 #[derive(Debug, Eq, PartialEq)]
+/// Error returned by [`Sender::try_send`] when a message cannot be queued immediately.
 pub enum TrySendError<T> {
+    /// The bounded queue is currently full.
     Full(T),
+    /// The receiver has been closed or dropped.
     Closed(T),
 }
 
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
+/// Error returned by [`Receiver::try_recv`] when no message is available immediately.
 pub enum TryRecvError {
+    /// The channel is still open, but currently empty.
     Empty,
+    /// The channel is closed and no more messages can arrive.
     Disconnected,
 }
 
@@ -215,12 +237,20 @@ impl<T: Send + 'static> Clone for UnboundedSender<T> {
 }
 
 impl<T: Send + 'static> Sender<T> {
+    /// Waits until the message can be queued.
+    ///
+    /// When the bounded channel is full, this future waits until the receiver frees capacity.
+    ///
+    /// # Panics
+    ///
+    /// Panics if this future is first polled outside a runtime-managed thread.
     pub async fn send(&self, value: T) -> Result<(), SendError<T>> {
         let mut value = Some(value);
         let mut wait = None;
         poll_fn(|cx| self.poll_send(cx, &mut value, &mut wait)).await
     }
 
+    /// Attempts to queue a message immediately.
     pub fn try_send(&self, value: T) -> Result<(), TrySendError<T>> {
         self.shared
             .lock()
@@ -228,6 +258,7 @@ impl<T: Send + 'static> Sender<T> {
             .try_send_now(value)
     }
 
+    /// Returns `true` if the receiver has been closed or dropped.
     pub fn is_closed(&self) -> bool {
         self.shared
             .lock()
@@ -306,6 +337,7 @@ impl<T: Send + 'static> Sender<T> {
 }
 
 impl<T: Send + 'static> UnboundedSender<T> {
+    /// Queues a message immediately.
     pub fn send(&self, value: T) -> Result<(), SendError<T>> {
         self.shared
             .lock()
@@ -316,6 +348,7 @@ impl<T: Send + 'static> UnboundedSender<T> {
             })
     }
 
+    /// Returns `true` if the receiver has been closed or dropped.
     pub fn is_closed(&self) -> bool {
         self.shared
             .lock()
@@ -325,11 +358,19 @@ impl<T: Send + 'static> UnboundedSender<T> {
 }
 
 impl<T: Send + 'static> Receiver<T> {
+    /// Waits for the next message.
+    ///
+    /// Returns `None` when the channel is closed and all buffered messages have been drained.
+    ///
+    /// # Panics
+    ///
+    /// Panics if this future is first polled outside a runtime-managed thread.
     pub async fn recv(&mut self) -> Option<T> {
         let mut wait = None;
         poll_fn(|cx| self.poll_recv(cx, &mut wait)).await
     }
 
+    /// Attempts to receive a message immediately.
     pub fn try_recv(&mut self) -> Result<T, TryRecvError> {
         let mut state = self
             .shared
@@ -345,6 +386,10 @@ impl<T: Send + 'static> Receiver<T> {
         }
     }
 
+    /// Closes the channel to future sends.
+    ///
+    /// Already-buffered messages remain available to [`recv`](Self::recv) and
+    /// [`try_recv`](Self::try_recv).
     pub fn close(&mut self) {
         self.shared
             .lock()
@@ -352,6 +397,7 @@ impl<T: Send + 'static> Receiver<T> {
             .close_receiver();
     }
 
+    /// Returns `true` if the channel is closed or all senders have been dropped.
     pub fn is_closed(&self) -> bool {
         let state = self
             .shared
diff --git a/lib/runtime/src/channel/oneshot.rs b/lib/runtime/src/channel/oneshot.rs
index 79ea4c4..631d7d9 100644
--- a/lib/runtime/src/channel/oneshot.rs
+++ b/lib/runtime/src/channel/oneshot.rs
@@ -1,3 +1,5 @@
+//! Single-use channels for handing one value from a sender to a receiver.
+
 use std::future::poll_fn;
 use std::pin::Pin;
 use std::sync::{Arc, Mutex};
@@ -6,6 +8,15 @@ use std::task::{Context, Poll};
 use crate::op::completion::{CompletionFuture, CompletionHandle};
 use crate::sys::linux::channel::runtime_waiter;
 
+/// Creates a single-use channel for transferring one value from a [`Sender`] to a [`Receiver`].
+///
+/// # Examples
+///
+/// ```
+/// let (sender, mut receiver) = ruin_runtime::channel::oneshot::channel::<usize>();
+/// sender.send(7).unwrap();
+/// assert_eq!(receiver.try_recv(), Ok(7));
+/// ```
 pub fn channel<T: Send + 'static>() -> (Sender<T>, Receiver<T>) {
     let shared = Arc::new(Mutex::new(State {
         value: None,
@@ -24,10 +35,12 @@ pub fn channel<T: Send + 'static>() -> (Sender<T>, Receiver<T>) {
     )
 }
 
+/// Sending half of a oneshot channel.
 pub struct Sender<T: Send + 'static> {
     shared: Option<Arc<Mutex<State<T>>>>,
 }
 
+/// Receiving half of a oneshot channel.
 pub struct Receiver<T: Send + 'static> {
     shared: Arc<Mutex<State<T>>>,
     consumed: bool,
@@ -41,18 +54,27 @@ struct State<T: Send + 'static> {
 }
 
 #[derive(Debug, Eq, PartialEq)]
+/// Error returned when a oneshot send fails because the receiver is gone or closed.
 pub struct SendError<T>(pub T);
 
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
+/// Error returned when a oneshot receive observes a closed channel with no value.
 pub struct RecvError;
 
 #[derive(Debug, Eq, PartialEq)]
+/// Non-blocking receive errors for [`Receiver::try_recv`].
 pub enum TryRecvError {
+    /// No value has been sent yet, and the sender is still alive.
     Empty,
+    /// The channel can never yield a value.
     Closed,
 }
 
 impl<T: Send + 'static> Sender<T> {
+    /// Sends `value` into the channel.
+    ///
+    /// This consumes the sender. If the receiver is already waiting on a runtime thread, the
+    /// receive future is woken through the runtime's cross-thread notifier path.
     pub fn send(mut self, value: T) -> Result<(), SendError<T>> {
         let Some(shared) = self.shared.take() else {
             return Err(SendError(value));
@@ -80,6 +102,7 @@ impl<T: Send + 'static> Sender<T> {
         Ok(())
     }
 
+    /// Returns `true` if the receiver has been closed or dropped.
     pub fn is_closed(&self) -> bool {
         self.shared.as_ref().is_none_or(|shared| {
             shared
@@ -91,11 +114,19 @@ impl<T: Send + 'static> Sender<T> {
 }
 
 impl<T: Send + 'static> Receiver<T> {
+    /// Waits for the channel's value.
+    ///
+    /// # Panics
+    ///
+    /// Panics if this future is first polled outside a runtime-managed thread. Async channel
+    /// waiting registers with the current runtime thread so it can be woken through the driver's
+    /// notification path.
     pub async fn recv(&mut self) -> Result<T, RecvError> {
         let mut wait = None;
         poll_fn(|cx| self.poll_recv(cx, &mut wait)).await
     }
 
+    /// Attempts to receive the value without waiting.
     pub fn try_recv(&mut self) -> Result<T, TryRecvError> {
         if self.consumed {
             return Err(TryRecvError::Closed);
@@ -118,6 +149,10 @@ impl<T: Send + 'static> Receiver<T> {
         }
     }
 
+    /// Closes the receiver.
+    ///
+    /// Closing prevents future sends from succeeding. If a value has already been sent, it can
+    /// still be retrieved.
     pub fn close(&mut self) {
         let mut state = self
             .shared
@@ -126,6 +161,7 @@ impl<T: Send + 'static> Receiver<T> {
         state.receiver_closed = true;
     }
 
+    /// Returns `true` if the channel is closed to future sends.
     pub fn is_closed(&self) -> bool {
         let state = self
             .shared
diff --git a/lib/runtime/src/fs.rs b/lib/runtime/src/fs.rs
index aa33f2d..be7edcf 100644
--- a/lib/runtime/src/fs.rs
+++ b/lib/runtime/src/fs.rs
@@ -4,6 +4,9 @@
 //! - Dropping an I/O future cancels interest in the result.
 //! - The runtime issues best-effort kernel cancellation where supported.
 //! - The underlying OS operation may still complete after the future is dropped.
+//!
+//! The public surface intentionally mirrors `std::fs` where that shape makes sense, while using
+//! async methods for operations that may block the caller.
 
 use std::ffi::OsStr;
 use std::io;
@@ -21,33 +24,43 @@ struct FileInner {
     fd: OwnedFd,
 }
 
+/// Async file handle.
+///
+/// `File` is cheap to clone internally and supports both cursor-based sequential I/O and
+/// offset-based positioned I/O.
 pub struct File {
     inner: Arc<FileInner>,
 }
 
+/// Builder used to configure how a [`File`] is opened.
 pub struct OpenOptions {
     inner: OpOpenOptions,
 }
 
 #[derive(Clone, Debug, Eq, PartialEq)]
+/// File metadata returned by [`metadata`] or [`File::metadata`].
 pub struct Metadata {
     inner: RawMetadata,
 }
 
+/// Async directory-entry stream returned by [`read_dir`].
 pub struct ReadDir {
     inner: sys_fs::ReadDirStream,
 }
 
 #[derive(Clone, Debug, Eq, PartialEq)]
+/// Directory entry yielded by [`ReadDir::next_entry`].
 pub struct DirEntry {
     inner: OpDirEntry,
 }
 
 impl File {
+    /// Opens an existing file for reading.
     pub async fn open(path: impl AsRef<Path>) -> io::Result<Self> {
         OpenOptions::new().read(true).open(path).await
     }
 
+    /// Opens a file for writing, creating or truncating it first.
     pub async fn create(path: impl AsRef<Path>) -> io::Result<Self> {
         OpenOptions::new()
             .write(true)
@@ -57,10 +70,12 @@ impl File {
             .await
     }
 
+    /// Reads bytes from the file's current cursor position.
     pub async fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
         self.read_impl(None, buf).await
     }
 
+    /// Reads exactly `buf.len()` bytes from the current cursor position.
     pub async fn read_exact(&mut self, mut buf: &mut [u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let read = self.read(buf).await?;
@@ -75,10 +90,12 @@ impl File {
         Ok(())
     }
 
+    /// Writes bytes at the file's current cursor position.
     pub async fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
         self.write_impl(None, buf).await
     }
 
+    /// Writes the entire buffer at the file's current cursor position.
     pub async fn write_all(&mut self, mut buf: &[u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let written = self.write(buf).await?;
@@ -93,22 +110,30 @@ impl File {
         Ok(())
     }
 
+    /// Flushes any userspace buffering associated with this handle.
+    ///
+    /// The current implementation does not add additional buffering beyond the kernel file
+    /// description, so this is effectively a no-op.
     pub async fn flush(&mut self) -> io::Result<()> {
         Ok(())
     }
 
+    /// Synchronizes file contents and metadata to stable storage.
     pub async fn sync_all(&self) -> io::Result<()> {
         sys_fs::sync_all(FsOp::SyncAll { fd: self.raw_fd() }).await
     }
 
+    /// Synchronizes file contents to stable storage.
     pub async fn sync_data(&self) -> io::Result<()> {
         sys_fs::sync_data(FsOp::SyncData { fd: self.raw_fd() }).await
     }
 
+    /// Reads bytes starting at `offset` without using the shared file cursor.
     pub async fn read_at(&self, offset: u64, buf: &mut [u8]) -> io::Result<usize> {
         self.read_impl(Some(offset), buf).await
     }
 
+    /// Reads exactly `buf.len()` bytes starting at `offset`.
     pub async fn read_exact_at(&self, mut offset: u64, mut buf: &mut [u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let read = self.read_at(offset, buf).await?;
@@ -124,10 +149,12 @@ impl File {
         Ok(())
     }
 
+    /// Writes bytes starting at `offset` without using the shared file cursor.
     pub async fn write_at(&self, offset: u64, buf: &[u8]) -> io::Result<usize> {
         self.write_impl(Some(offset), buf).await
     }
 
+    /// Writes the entire buffer starting at `offset`.
     pub async fn write_all_at(&self, mut offset: u64, mut buf: &[u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let written = self.write_at(offset, buf).await?;
@@ -143,6 +170,7 @@ impl File {
         Ok(())
     }
 
+    /// Returns metadata for this file handle.
     pub async fn metadata(&self) -> io::Result<Metadata> {
         sys_fs::metadata(FsOp::Metadata {
             target: MetadataTarget::File(self.raw_fd()),
@@ -152,6 +180,7 @@ impl File {
         .map(Metadata::from_raw)
     }
 
+    /// Truncates or extends the underlying file to `len` bytes.
     pub async fn set_len(&self, len: u64) -> io::Result<()> {
         sys_fs::set_len(FsOp::SetLen {
             fd: self.raw_fd(),
@@ -160,6 +189,9 @@ impl File {
         .await
     }
 
+    /// Duplicates the underlying file description.
+    ///
+    /// As with `std::fs::File::try_clone`, the cloned handle shares kernel-managed cursor state.
     pub async fn try_clone(&self) -> io::Result<Self> {
         sys_fs::try_clone(FsOp::Duplicate { fd: self.raw_fd() })
             .await
@@ -200,42 +232,63 @@ impl File {
 }
 
 impl OpenOptions {
+    /// Creates a blank set of open options.
     pub fn new() -> Self {
         Self {
             inner: OpOpenOptions::default(),
         }
     }
 
+    /// Controls read access.
     pub fn read(&mut self, value: bool) -> &mut Self {
         self.inner.read = value;
         self
     }
 
+    /// Controls write access.
     pub fn write(&mut self, value: bool) -> &mut Self {
         self.inner.write = value;
         self
     }
 
+    /// Controls append mode.
     pub fn append(&mut self, value: bool) -> &mut Self {
         self.inner.append = value;
         self
     }
 
+    /// Controls whether the file is truncated after opening.
     pub fn truncate(&mut self, value: bool) -> &mut Self {
         self.inner.truncate = value;
         self
     }
 
+    /// Controls whether the file is created if it does not already exist.
     pub fn create(&mut self, value: bool) -> &mut Self {
         self.inner.create = value;
         self
     }
 
+    /// Controls whether opening must create a brand-new file.
     pub fn create_new(&mut self, value: bool) -> &mut Self {
         self.inner.create_new = value;
         self
     }
 
+    /// Opens a file with the configured options.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// # let _ = || async {
+    /// let file = ruin_runtime::fs::OpenOptions::new()
+    ///     .read(true)
+    ///     .write(true)
+    ///     .open("example.txt")
+    ///     .await;
+    /// # let _ = file;
+    /// # };
+    /// ```
     pub async fn open(&self, path: impl AsRef<Path>) -> io::Result<File> {
         sys_fs::open(FsOp::Open {
             path: path.as_ref().to_path_buf(),
@@ -257,32 +310,39 @@ impl Metadata {
         Self { inner }
     }
 
+    /// Returns the file length in bytes.
     pub fn len(&self) -> u64 {
         self.inner.len
     }
 
+    /// Returns `true` if the file length is zero.
     pub fn is_empty(&self) -> bool {
         self.len() == 0
     }
 
+    /// Returns `true` if this metadata describes a regular file.
     pub fn is_file(&self) -> bool {
         self.inner.file_type == RawFileType::File
     }
 
+    /// Returns `true` if this metadata describes a directory.
     pub fn is_dir(&self) -> bool {
         self.inner.file_type == RawFileType::Directory
     }
 
+    /// Returns `true` if this metadata describes a symbolic link.
     pub fn is_symlink(&self) -> bool {
         self.inner.file_type == RawFileType::Symlink
     }
 
+    /// Returns the raw POSIX mode bits reported by the platform backend.
     pub fn mode(&self) -> u16 {
         self.inner.mode
     }
 }
 
 impl ReadDir {
+    /// Returns the next directory entry, or `None` once the stream is exhausted.
     pub async fn next_entry(&mut self) -> io::Result<Option<DirEntry>> {
         self.inner
             .next_entry()
@@ -292,19 +352,23 @@ impl ReadDir {
 }
 
 impl DirEntry {
+    /// Returns the full path to this directory entry.
     pub fn path(&self) -> PathBuf {
         self.inner.path.clone()
     }
 
+    /// Returns the file name portion of this directory entry.
     pub fn file_name(&self) -> &OsStr {
         self.inner.file_name.as_os_str()
     }
 
+    /// Resolves metadata for this entry.
     pub async fn metadata(&self) -> io::Result<Metadata> {
         metadata(self.path()).await
     }
 }
 
+/// Reads the entire contents of a file into memory.
 pub async fn read(path: impl AsRef<Path>) -> io::Result<Vec<u8>> {
     let mut file = File::open(path.as_ref()).await?;
     let mut output = Vec::new();
@@ -319,11 +383,13 @@ pub async fn read(path: impl AsRef<Path>) -> io::Result<Vec<u8>> {
     }
 }
 
+/// Reads the entire contents of a UTF-8 file into a [`String`].
 pub async fn read_to_string(path: impl AsRef<Path>) -> io::Result<String> {
     let bytes = read(path).await?;
     String::from_utf8(bytes).map_err(|error| io::Error::new(io::ErrorKind::InvalidData, error))
 }
 
+/// Replaces the contents of a file with `data`, creating it if needed.
 pub async fn write(path: impl AsRef<Path>, data: impl AsRef<[u8]>) -> io::Result<()> {
     let mut file = OpenOptions::new()
         .write(true)
@@ -334,6 +400,7 @@ pub async fn write(path: impl AsRef<Path>, data: impl AsRef<[u8]>) -> io::Result
     file.write_all(data.as_ref()).await
 }
 
+/// Returns metadata for a filesystem path.
 pub async fn metadata(path: impl AsRef<Path>) -> io::Result<Metadata> {
     sys_fs::metadata(FsOp::Metadata {
         target: MetadataTarget::Path(path.as_ref().to_path_buf()),
@@ -343,6 +410,7 @@ pub async fn metadata(path: impl AsRef<Path>) -> io::Result<Metadata> {
     .map(Metadata::from_raw)
 }
 
+/// Creates a single directory.
 pub async fn create_dir(path: impl AsRef<Path>) -> io::Result<()> {
     sys_fs::create_dir(FsOp::CreateDir {
         path: path.as_ref().to_path_buf(),
@@ -352,6 +420,7 @@ pub async fn create_dir(path: impl AsRef<Path>) -> io::Result<()> {
     .await
 }
 
+/// Creates a directory and any missing parent directories.
 pub async fn create_dir_all(path: impl AsRef<Path>) -> io::Result<()> {
     let path = path.as_ref();
     let mut current = PathBuf::new();
@@ -372,6 +441,7 @@ pub async fn create_dir_all(path: impl AsRef<Path>) -> io::Result<()> {
     Ok(())
 }
 
+/// Removes a file.
 pub async fn remove_file(path: impl AsRef<Path>) -> io::Result<()> {
     sys_fs::remove_file(FsOp::RemoveFile {
         path: path.as_ref().to_path_buf(),
@@ -379,6 +449,7 @@ pub async fn remove_file(path: impl AsRef<Path>) -> io::Result<()> {
     .await
 }
 
+/// Removes an empty directory.
 pub async fn remove_dir(path: impl AsRef<Path>) -> io::Result<()> {
     sys_fs::remove_dir(FsOp::RemoveDir {
         path: path.as_ref().to_path_buf(),
@@ -386,6 +457,7 @@ pub async fn remove_dir(path: impl AsRef<Path>) -> io::Result<()> {
     .await
 }
 
+/// Renames or moves a filesystem entry.
 pub async fn rename(from: impl AsRef<Path>, to: impl AsRef<Path>) -> io::Result<()> {
     sys_fs::rename(FsOp::Rename {
         from: from.as_ref().to_path_buf(),
@@ -394,6 +466,16 @@ pub async fn rename(from: impl AsRef<Path>, to: impl AsRef<Path>) -> io::Result<
     .await
 }
 
+/// Opens an async directory-entry stream.
+///
+/// # Examples
+///
+/// ```
+/// # let _ = || async {
+/// let mut entries = ruin_runtime::fs::read_dir(".").await.unwrap();
+/// let _ = entries.next_entry().await;
+/// # };
+/// ```
 pub async fn read_dir(path: impl AsRef<Path>) -> io::Result<ReadDir> {
     sys_fs::read_dir(FsOp::ReadDir {
         path: path.as_ref().to_path_buf(),
diff --git a/lib/runtime/src/lib.rs b/lib/runtime/src/lib.rs
index 90390ff..9559c9d 100644
--- a/lib/runtime/src/lib.rs
+++ b/lib/runtime/src/lib.rs
@@ -1,6 +1,21 @@
+//! Runtime, driver, async I/O, and channel primitives for RUIN.
+//!
+//! The crate is centered around a single-threaded event loop with explicit worker threads,
+//! JavaScript-style microtask/macrotask scheduling, and Linux `io_uring`-backed I/O.
+//!
+//! Most users will start with:
+//!
+//! - [`main`] or [`async_main`] for executable entry points
+//! - [`run`], [`queue_task`], [`queue_microtask`], and [`queue_future`] for event-loop work
+//! - [`fs`], [`net`], [`time`], and [`channel`] for async runtime services
+//!
+//! # Platform support
+//!
+//! `ruin-runtime` currently targets Linux on `x86_64`.
+//!
 //! RUIN runtime foundations.
 //!
-//! This crate provides a Linux x86_64 runtime substrate: the mesh allocator, the reactor, and a
+//! This crate provides a Linux x86_64 runtime substrate: the mesh allocator, the driver, and a
 //! single-threaded runtime loop with worker-thread task forwarding.
 
 #![feature(thread_local)]
@@ -13,13 +28,31 @@ extern crate alloc;
 pub mod channel;
 pub mod fs;
 pub mod net;
+#[doc(hidden)]
 pub mod op;
+#[doc(hidden)]
 pub mod platform;
+#[doc(hidden)]
 pub mod sys;
 pub mod time;
 
-pub use ruin_runtime_proc_macros::{async_main, main};
+/// Marks an `async fn main()` as the runtime entry point.
+///
+/// The macro generates a real Rust `main` that queues the returned future onto the main runtime
+/// thread before calling [`run`].
+pub use ruin_runtime_proc_macros::async_main;
+/// Marks a synchronous `fn main()` as the runtime entry point.
+///
+/// The macro generates a real Rust `main` that queues the function body onto the main runtime
+/// thread before calling [`run`].
+pub use ruin_runtime_proc_macros::main;
 
+/// Driver primitives re-exported from the Linux x86_64 backend.
+#[cfg(all(target_os = "linux", target_arch = "x86_64"))]
+pub use platform::linux_x86_64::driver::{
+    Driver, ReadyEvents, ThreadNotifier, create, create_driver, monotonic_now,
+};
+/// Public mesh-allocator surface re-exported from the Linux x86_64 backend.
 #[cfg(all(target_os = "linux", target_arch = "x86_64"))]
 pub use platform::linux_x86_64::mesh_alloc::{
     ActiveMeshGuard, Arena, AtomicBitmap, BitIter, CLASS_TO_SIZE, CompactionAdvice,
@@ -32,12 +65,10 @@ pub use platform::linux_x86_64::mesh_alloc::{
     page_shift, page_size, retry_on_efault, retry_on_efault_ptrs, round_up_to_page,
     runtime_slots_per_span, size_class_for,
 };
+/// Additional allocator helpers re-exported from the Linux x86_64 backend.
 #[cfg(all(target_os = "linux", target_arch = "x86_64"))]
 pub use platform::linux_x86_64::mesh_alloc::{FreelistId, bitmaps_meshable};
-#[cfg(all(target_os = "linux", target_arch = "x86_64"))]
-pub use platform::linux_x86_64::reactor::{
-    Reactor, ReadyEvents, ThreadNotifier, create, create_reactor, monotonic_now,
-};
+/// Runtime/event-loop primitives re-exported from the Linux x86_64 backend.
 #[cfg(all(target_os = "linux", target_arch = "x86_64"))]
 pub use platform::linux_x86_64::runtime::{
     IntervalHandle, JoinHandle, ThreadHandle, TimeoutHandle, WorkerHandle, clear_interval,
@@ -45,6 +76,9 @@ pub use platform::linux_x86_64::runtime::{
     set_interval, set_timeout, spawn_worker, yield_now,
 };
 
+/// Returns the default global mesh allocator configuration for this crate.
+///
+/// This is useful when embedding the allocator in a `#[global_allocator]` static.
 pub const fn default_global_allocator() -> GlobalMeshAllocator {
     GlobalMeshAllocator::with_default_config()
 }
diff --git a/lib/runtime/src/net.rs b/lib/runtime/src/net.rs
index aa9a961..7902f53 100644
--- a/lib/runtime/src/net.rs
+++ b/lib/runtime/src/net.rs
@@ -1,4 +1,7 @@
 //! Portable async networking API.
+//!
+//! The public surface follows the general shape of `std::net`, but uses async methods for socket
+//! operations that would otherwise block the caller.
 
 use std::future::Future;
 use std::io;
@@ -40,6 +43,10 @@ type PendingRead = Pin<Box<dyn Future<Output = io::Result<Vec<u8>>> + 'static>>;
 type PendingWrite = Pin<Box<dyn Future<Output = io::Result<usize>> + 'static>>;
 type PendingShutdown = Pin<Box<dyn Future<Output = io::Result<()>> + 'static>>;
 
+/// Async TCP stream.
+///
+/// This type also implements Hyper's runtime I/O traits, allowing it to be used directly as an
+/// HTTP transport.
 pub struct TcpStream {
     inner: Arc<TcpStreamInner>,
     pending_read: Option<PendingRead>,
@@ -48,16 +55,19 @@ pub struct TcpStream {
 }
 
 #[derive(Clone, Debug)]
+/// Async TCP listening socket.
 pub struct TcpListener {
     inner: Arc<TcpListenerInner>,
 }
 
 #[derive(Debug)]
+/// Async UDP socket.
 pub struct UdpSocket {
     inner: Arc<UdpSocketInner>,
 }
 
 impl TcpStream {
+    /// Connects to the first resolved address that succeeds.
     pub async fn connect<A>(addr: A) -> io::Result<Self>
     where
         A: ToSocketAddrs + Send + 'static,
@@ -79,6 +89,7 @@ impl TcpStream {
         }))
     }
 
+    /// Connects to `addr`, failing if the deadline elapses first.
     pub async fn connect_timeout(addr: &SocketAddr, timeout: Duration) -> io::Result<Self> {
         validate_timeout(timeout)?;
         crate::sys::linux::net::connect_stream_timeout(*addr, timeout)
@@ -86,6 +97,7 @@ impl TcpStream {
             .map(Self::from_owned_fd)
     }
 
+    /// Reads bytes from the stream.
     pub async fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
         let data = match self.read_timeout_value() {
             Some(timeout) => {
@@ -105,6 +117,7 @@ impl TcpStream {
         Ok(read)
     }
 
+    /// Reads exactly `buf.len()` bytes from the stream.
     pub async fn read_exact(&mut self, mut buf: &mut [u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let read = self.read(buf).await?;
@@ -119,6 +132,7 @@ impl TcpStream {
         Ok(())
     }
 
+    /// Writes bytes to the stream.
     pub async fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
         match self.write_timeout_value() {
             Some(timeout) => {
@@ -135,6 +149,7 @@ impl TcpStream {
         }
     }
 
+    /// Writes the entire buffer to the stream.
     pub async fn write_all(&mut self, mut buf: &[u8]) -> io::Result<()> {
         while !buf.is_empty() {
             let written = self.write(buf).await?;
@@ -149,6 +164,7 @@ impl TcpStream {
         Ok(())
     }
 
+    /// Shuts down the read, write, or both halves of the connection.
     pub async fn shutdown(&self, how: Shutdown) -> io::Result<()> {
         crate::sys::linux::net::shutdown(NetOp::Shutdown {
             fd: self.raw_fd(),
@@ -157,50 +173,65 @@ impl TcpStream {
         .await
     }
 
+    /// Duplicates the underlying stream socket.
     pub async fn try_clone(&self) -> io::Result<Self> {
         crate::sys::linux::net::duplicate(self.raw_fd())
             .await
             .map(Self::from_owned_fd)
     }
 
+    /// Returns the local socket address of this stream.
     pub fn local_addr(&self) -> io::Result<SocketAddr> {
         crate::sys::linux::net::local_addr(self.raw_fd())
     }
 
+    /// Returns the remote peer address of this stream.
     pub fn peer_addr(&self) -> io::Result<SocketAddr> {
         crate::sys::linux::net::peer_addr(self.raw_fd())
     }
 
+    /// Reads the current `TCP_NODELAY` setting.
     pub fn nodelay(&self) -> io::Result<bool> {
         crate::sys::linux::net::nodelay(self.raw_fd())
     }
 
+    /// Enables or disables `TCP_NODELAY`.
     pub fn set_nodelay(&self, enabled: bool) -> io::Result<()> {
         crate::sys::linux::net::set_nodelay(self.raw_fd(), enabled)
     }
 
+    /// Reads the socket's IP time-to-live value.
     pub fn ttl(&self) -> io::Result<u32> {
         crate::sys::linux::net::ttl(self.raw_fd())
     }
 
+    /// Sets the socket's IP time-to-live value.
     pub fn set_ttl(&self, ttl: u32) -> io::Result<()> {
         crate::sys::linux::net::set_ttl(self.raw_fd(), ttl)
     }
 
+    /// Returns the read timeout used by async read operations on this handle.
     pub fn read_timeout(&self) -> io::Result<Option<Duration>> {
         Ok(self.read_timeout_value())
     }
 
+    /// Sets the read timeout used by async read operations on this handle.
+    ///
+    /// Passing `Some(Duration::ZERO)` is rejected.
     pub fn set_read_timeout(&self, timeout: Option<Duration>) -> io::Result<()> {
         validate_optional_timeout(timeout)?;
         self.inner.timeouts.lock().unwrap().read = timeout;
         Ok(())
     }
 
+    /// Returns the write timeout used by async write operations on this handle.
     pub fn write_timeout(&self) -> io::Result<Option<Duration>> {
         Ok(self.write_timeout_value())
     }
 
+    /// Sets the write timeout used by async write operations on this handle.
+    ///
+    /// Passing `Some(Duration::ZERO)` is rejected.
     pub fn set_write_timeout(&self, timeout: Option<Duration>) -> io::Result<()> {
         validate_optional_timeout(timeout)?;
         self.inner.timeouts.lock().unwrap().write = timeout;
@@ -233,6 +264,7 @@ impl TcpStream {
 }
 
 impl TcpListener {
+    /// Binds a TCP listener to the first resolved address that succeeds.
     pub async fn bind<A>(addr: A) -> io::Result<Self>
     where
         A: ToSocketAddrs + Send + 'static,
@@ -254,6 +286,7 @@ impl TcpListener {
         }))
     }
 
+    /// Accepts an incoming connection.
     pub async fn accept(&self) -> io::Result<(TcpStream, SocketAddr)> {
         let accepted = crate::sys::linux::net::accept(NetOp::Accept { fd: self.raw_fd() }).await?;
 
@@ -261,14 +294,17 @@ impl TcpListener {
         Ok((stream, accepted.peer_addr))
     }
 
+    /// Returns the local socket address of this listener.
     pub fn local_addr(&self) -> io::Result<SocketAddr> {
         crate::sys::linux::net::local_addr(self.raw_fd())
     }
 
+    /// Reads the listener socket's IP time-to-live value.
     pub fn ttl(&self) -> io::Result<u32> {
         crate::sys::linux::net::ttl(self.raw_fd())
     }
 
+    /// Sets the listener socket's IP time-to-live value.
     pub fn set_ttl(&self, ttl: u32) -> io::Result<()> {
         crate::sys::linux::net::set_ttl(self.raw_fd(), ttl)
     }
@@ -285,6 +321,7 @@ impl TcpListener {
 }
 
 impl UdpSocket {
+    /// Binds a UDP socket to the first resolved address that succeeds.
     pub async fn bind<A>(addr: A) -> io::Result<Self>
     where
         A: ToSocketAddrs + Send + 'static,
@@ -306,6 +343,10 @@ impl UdpSocket {
         }))
     }
 
+    /// Connects the socket to a default peer.
+    ///
+    /// Once connected, [`send`](Self::send), [`recv`](Self::recv), and [`peer_addr`](Self::peer_addr)
+    /// operate relative to that peer.
     pub async fn connect<A>(&self, addr: A) -> io::Result<()>
     where
         A: ToSocketAddrs + Send + 'static,
@@ -332,6 +373,7 @@ impl UdpSocket {
         }))
     }
 
+    /// Sends a datagram to the connected peer.
     pub async fn send(&self, buf: &[u8]) -> io::Result<usize> {
         match self.write_timeout_value() {
             Some(timeout) => {
@@ -348,6 +390,7 @@ impl UdpSocket {
         }
     }
 
+    /// Receives a datagram from the connected peer.
     pub async fn recv(&self, buf: &mut [u8]) -> io::Result<usize> {
         let data = match self.read_timeout_value() {
             Some(timeout) => {
@@ -367,6 +410,7 @@ impl UdpSocket {
         Ok(read)
     }
 
+    /// Peeks at the next datagram from the connected peer without consuming it.
     pub async fn peek(&self, buf: &mut [u8]) -> io::Result<usize> {
         let data = match self.read_timeout_value() {
             Some(timeout) => {
@@ -392,6 +436,7 @@ impl UdpSocket {
         Ok(read)
     }
 
+    /// Sends a datagram to `addr`.
     pub async fn send_to<A>(&self, buf: &[u8], addr: A) -> io::Result<usize>
     where
         A: ToSocketAddrs + Send + 'static,
@@ -435,6 +480,7 @@ impl UdpSocket {
         }))
     }
 
+    /// Receives a datagram and returns the sender address.
     pub async fn recv_from(&self, buf: &mut [u8]) -> io::Result<(usize, SocketAddr)> {
         let datagram = match self.read_timeout_value() {
             Some(timeout) => {
@@ -455,6 +501,7 @@ impl UdpSocket {
         Ok((read, datagram.peer_addr))
     }
 
+    /// Peeks at the next datagram and returns the sender address without consuming it.
     pub async fn peek_from(&self, buf: &mut [u8]) -> io::Result<(usize, SocketAddr)> {
         let datagram = match self.read_timeout_value() {
             Some(timeout) => {
@@ -480,50 +527,65 @@ impl UdpSocket {
         Ok((read, datagram.peer_addr))
     }
 
+    /// Duplicates the underlying UDP socket.
     pub async fn try_clone(&self) -> io::Result<Self> {
         crate::sys::linux::net::duplicate(self.raw_fd())
             .await
             .map(Self::from_owned_fd)
     }
 
+    /// Returns the local socket address of this socket.
     pub fn local_addr(&self) -> io::Result<SocketAddr> {
         crate::sys::linux::net::local_addr(self.raw_fd())
     }
 
+    /// Returns the connected peer address, if the socket has been connected.
     pub fn peer_addr(&self) -> io::Result<SocketAddr> {
         crate::sys::linux::net::peer_addr(self.raw_fd())
     }
 
+    /// Reads the `SO_BROADCAST` setting.
     pub fn broadcast(&self) -> io::Result<bool> {
         crate::sys::linux::net::broadcast(self.raw_fd())
     }
 
+    /// Enables or disables `SO_BROADCAST`.
     pub fn set_broadcast(&self, enabled: bool) -> io::Result<()> {
         crate::sys::linux::net::set_broadcast(self.raw_fd(), enabled)
     }
 
+    /// Reads the socket's IP time-to-live value.
     pub fn ttl(&self) -> io::Result<u32> {
         crate::sys::linux::net::ttl(self.raw_fd())
     }
 
+    /// Sets the socket's IP time-to-live value.
     pub fn set_ttl(&self, ttl: u32) -> io::Result<()> {
         crate::sys::linux::net::set_ttl(self.raw_fd(), ttl)
     }
 
+    /// Returns the read timeout used by async receive operations on this handle.
     pub fn read_timeout(&self) -> io::Result<Option<Duration>> {
         Ok(self.read_timeout_value())
     }
 
+    /// Sets the read timeout used by async receive operations on this handle.
+    ///
+    /// Passing `Some(Duration::ZERO)` is rejected.
     pub fn set_read_timeout(&self, timeout: Option<Duration>) -> io::Result<()> {
         validate_optional_timeout(timeout)?;
         self.inner.timeouts.lock().unwrap().read = timeout;
         Ok(())
     }
 
+    /// Returns the write timeout used by async send operations on this handle.
     pub fn write_timeout(&self) -> io::Result<Option<Duration>> {
         Ok(self.write_timeout_value())
     }
 
+    /// Sets the write timeout used by async send operations on this handle.
+    ///
+    /// Passing `Some(Duration::ZERO)` is rejected.
     pub fn set_write_timeout(&self, timeout: Option<Duration>) -> io::Result<()> {
         validate_optional_timeout(timeout)?;
         self.inner.timeouts.lock().unwrap().write = timeout;
diff --git a/lib/runtime/src/platform/linux_x86_64/reactor.rs b/lib/runtime/src/platform/linux_x86_64/driver.rs
similarity index 83%
rename from lib/runtime/src/platform/linux_x86_64/reactor.rs
rename to lib/runtime/src/platform/linux_x86_64/driver.rs
index 3a833cb..f4578bd 100644
--- a/lib/runtime/src/platform/linux_x86_64/reactor.rs
+++ b/lib/runtime/src/platform/linux_x86_64/driver.rs
@@ -1,3 +1,5 @@
+//! Public runtime driver primitives.
+
 use std::cell::Cell;
 use std::cell::RefCell;
 use std::collections::BTreeMap;
@@ -51,23 +53,29 @@ impl NotifierInner {
 }
 
 #[derive(Clone)]
+/// Cross-thread notifier for a runtime thread's driver.
 pub struct ThreadNotifier {
     inner: Arc<NotifierInner>,
 }
 
 impl ThreadNotifier {
+    /// Sends a wake notification to the target runtime thread.
     pub fn notify(&self) -> io::Result<()> {
         self.inner.notify()
     }
 }
 
 #[derive(Debug, Default, Clone, Copy, Eq, PartialEq)]
+/// Readiness information returned by [`Driver::poll`].
 pub struct ReadyEvents {
+    /// One or more timer expirations are pending.
     pub timer: bool,
+    /// One or more cross-thread wake notifications are pending.
     pub wake: bool,
 }
 
-pub struct Reactor {
+/// Low-level Linux runtime driver backed by `io_uring`.
+pub struct Driver {
     ring: IoUring,
     notifier: Arc<NotifierInner>,
     next_token: Cell<u64>,
@@ -77,11 +85,16 @@ pub struct Reactor {
     completions: RefCell<BTreeMap<u64, CompletionHandler>>,
 }
 
-pub fn create() -> io::Result<(Reactor, ThreadNotifier)> {
-    create_reactor()
+/// Creates a new driver and its paired [`ThreadNotifier`].
+pub fn create() -> io::Result<(Driver, ThreadNotifier)> {
+    create_driver()
 }
 
-pub fn create_reactor() -> io::Result<(Reactor, ThreadNotifier)> {
+/// Creates a new driver and its paired [`ThreadNotifier`].
+///
+/// This is identical to [`create`] and exists as a more explicit name for callers that want to
+/// emphasize driver construction.
+pub fn create_driver() -> io::Result<(Driver, ThreadNotifier)> {
     let ring = IoUring::new(64)?;
     let notifier = Arc::new(NotifierInner {
         ring_fd: ring.ring_fd(),
@@ -89,7 +102,7 @@ pub fn create_reactor() -> io::Result<(Reactor, ThreadNotifier)> {
     });
 
     Ok((
-        Reactor {
+        Driver {
             ring,
             notifier: Arc::clone(&notifier),
             next_token: Cell::new(1),
@@ -102,7 +115,7 @@ pub fn create_reactor() -> io::Result<(Reactor, ThreadNotifier)> {
     ))
 }
 
-impl Reactor {
+impl Driver {
     pub(crate) fn bind_current_thread(&self) {
         self.ring.bind_current_thread();
     }
@@ -111,6 +124,7 @@ impl Reactor {
         self.ring.unbind_current_thread();
     }
 
+    /// Polls the driver without blocking.
     pub fn poll(&self) -> io::Result<Option<ReadyEvents>> {
         let mut ready = ReadyEvents::default();
         let saw_any = self
@@ -119,10 +133,14 @@ impl Reactor {
         if saw_any { Ok(Some(ready)) } else { Ok(None) }
     }
 
+    /// Blocks until at least one completion is available.
     pub fn wait(&self) -> io::Result<()> {
         self.ring.wait_for_cqe()
     }
 
+    /// Updates the currently armed timer deadline.
+    ///
+    /// Passing `None` removes any active timer.
     pub fn rearm_timer(&self, deadline: Option<Duration>) -> io::Result<()> {
         match (self.active_timer_token.get(), deadline) {
             (Some(active), Some(deadline)) => {
@@ -171,6 +189,7 @@ impl Reactor {
             })
     }
 
+    /// Drains the accumulated wake notification count.
     pub fn drain_wake(&self) -> io::Result<u64> {
         let wakes = self.pending_wakes.replace(0);
         if wakes == 0 {
@@ -183,6 +202,7 @@ impl Reactor {
         }
     }
 
+    /// Drains the accumulated timer-expiration count.
     pub fn drain_timer(&self) -> io::Result<u64> {
         let timers = self.pending_timers.replace(0);
         if timers == 0 {
@@ -234,12 +254,13 @@ impl Reactor {
     }
 }
 
-impl Drop for Reactor {
+impl Drop for Driver {
     fn drop(&mut self) {
         self.notifier.closed.store(true, Ordering::Release);
     }
 }
 
+/// Returns the current monotonic time used by the runtime timer system.
 pub fn monotonic_now() -> io::Result<Duration> {
     let mut now = std::mem::MaybeUninit::<libc::timespec>::uninit();
     let result = unsafe { libc::clock_gettime(libc::CLOCK_MONOTONIC, now.as_mut_ptr()) };
@@ -268,16 +289,16 @@ fn decode_token_kind(token: u64) -> Option<CompletionKind> {
 
 #[cfg(test)]
 mod tests {
-    use super::{create_reactor, monotonic_now};
+    use super::{create_driver, monotonic_now};
     use std::thread;
     use std::time::Duration;
 
     #[test]
     fn notifier_wakes_target_ring() {
-        let (sender, _) = create_reactor().expect("sender reactor should initialize");
+        let (sender, _) = create_driver().expect("sender driver should initialize");
         sender.bind_current_thread();
 
-        let (target, notifier) = create_reactor().expect("target reactor should initialize");
+        let (target, notifier) = create_driver().expect("target driver should initialize");
         notifier.notify().expect("notify should succeed");
 
         let ready = loop {
@@ -295,7 +316,7 @@ mod tests {
 
     #[test]
     fn notifier_wakes_target_ring_from_plain_thread() {
-        let (target, notifier) = create_reactor().expect("target reactor should initialize");
+        let (target, notifier) = create_driver().expect("target driver should initialize");
 
         thread::spawn(move || {
             notifier.notify().expect("notify should succeed");
@@ -317,14 +338,14 @@ mod tests {
 
     #[test]
     fn timeout_reports_deadlines() {
-        let (reactor, _notifier) = create_reactor().expect("reactor should initialize");
+        let (driver, _notifier) = create_driver().expect("driver should initialize");
         let deadline = monotonic_now().expect("clock should work") + Duration::from_millis(20);
-        reactor
+        driver
             .rearm_timer(Some(deadline))
             .expect("timer should arm");
 
         let ready = loop {
-            if let Some(ready) = reactor.poll().expect("poll should succeed") {
+            if let Some(ready) = driver.poll().expect("poll should succeed") {
                 break ready;
             }
             thread::sleep(Duration::from_millis(5));
@@ -332,9 +353,6 @@ mod tests {
 
         assert!(ready.timer);
         assert!(!ready.wake);
-        assert_eq!(
-            reactor.drain_timer().expect("timer drain should succeed"),
-            1
-        );
+        assert_eq!(driver.drain_timer().expect("timer drain should succeed"), 1);
     }
 }
diff --git a/lib/runtime/src/platform/linux_x86_64/mod.rs b/lib/runtime/src/platform/linux_x86_64/mod.rs
index a6a1110..706a729 100644
--- a/lib/runtime/src/platform/linux_x86_64/mod.rs
+++ b/lib/runtime/src/platform/linux_x86_64/mod.rs
@@ -1,4 +1,4 @@
+pub mod driver;
 pub mod mesh_alloc;
-pub mod reactor;
 pub mod runtime;
 pub(crate) mod uring;
diff --git a/lib/runtime/src/platform/linux_x86_64/runtime.rs b/lib/runtime/src/platform/linux_x86_64/runtime.rs
index 5deffd2..ef2ba5a 100644
--- a/lib/runtime/src/platform/linux_x86_64/runtime.rs
+++ b/lib/runtime/src/platform/linux_x86_64/runtime.rs
@@ -1,3 +1,5 @@
+//! Public runtime loop and worker-thread primitives.
+
 use std::cell::{Cell, RefCell};
 use std::collections::{BTreeMap, VecDeque};
 use std::future::Future;
@@ -9,7 +11,7 @@ use std::sync::{Arc, Mutex, MutexGuard};
 use std::task::{Context, Poll, RawWaker, RawWakerVTable, Waker};
 use std::time::Duration;
 
-use super::reactor::{Reactor, ThreadNotifier, create as create_reactor, monotonic_now};
+use super::driver::{Driver, ThreadNotifier, create_driver, monotonic_now};
 
 type LocalTask = Box<dyn FnOnce() + 'static>;
 type SendTask = Box<dyn FnOnce() + Send + 'static>;
@@ -19,16 +21,19 @@ type LocalBoxFuture = Pin<Box<dyn Future<Output = ()> + 'static>>;
 static mut CURRENT_THREAD: *mut ThreadState = ptr::null_mut();
 
 #[derive(Clone)]
+/// Handle for queueing work onto a specific runtime thread.
 pub struct ThreadHandle {
     shared: Arc<ThreadShared>,
 }
 
+/// Handle for a worker runtime thread spawned with [`spawn_worker`].
 pub struct WorkerHandle {
     thread: ThreadHandle,
     completion: Arc<WorkerCompletion>,
 }
 
 #[derive(Clone)]
+/// Handle returned by [`set_timeout`].
 pub struct TimeoutHandle {
     id: usize,
     owner: *const ThreadState,
@@ -36,20 +41,37 @@ pub struct TimeoutHandle {
 }
 
 #[derive(Clone)]
+/// Handle returned by [`set_interval`].
 pub struct IntervalHandle {
     id: usize,
     owner: *const ThreadState,
     _local: Rc<()>,
 }
 
+/// Handle returned by [`queue_future`].
+///
+/// Awaiting a join handle yields the output of the queued future.
 pub struct JoinHandle<T> {
     state: Rc<JoinState<T>>,
 }
 
+/// Future returned by [`yield_now`].
+///
+/// Awaiting this future will immediately yield control back to the runtime scheduler, allowing other queued microtasks
+/// to run before the current task continues executing. Note that continuation of futures runs as a microtask, so this
+/// can only yield to other microtasks and not to macrotasks (driver events such as file or network I/O, timers, or
+/// channel messages).
 pub struct YieldNow {
     yielded: bool,
 }
 
+/// Returns a handle for the current runtime thread.
+///
+/// If the current thread has not yet entered the runtime, the runtime state is initialized lazily.
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its driver for the current thread.
 pub fn current_thread_handle() -> ThreadHandle {
     current_thread().handle()
 }
@@ -58,10 +80,17 @@ pub(crate) fn try_current_thread_handle() -> Option<ThreadHandle> {
     unsafe { (!CURRENT_THREAD.is_null()).then(|| (*CURRENT_THREAD).handle()) }
 }
 
-pub(crate) fn with_current_reactor<T>(f: impl FnOnce(&Reactor) -> T) -> T {
-    f(&current_thread().reactor)
+pub(crate) fn with_current_driver<T>(f: impl FnOnce(&Driver) -> T) -> T {
+    f(&current_thread().driver)
 }
 
+/// Queues a macrotask on the current runtime thread.
+///
+/// The task runs after all currently-queued macrotasks, and after all microtasks.
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its state for the current thread.
 pub fn queue_task<F>(task: F)
 where
     F: FnOnce() + 'static,
@@ -69,6 +98,14 @@ where
     push_local_macrotask(Box::new(task));
 }
 
+/// Queues a microtask on the current runtime thread.
+///
+/// Microtasks run before the next macrotask turn, mirroring JavaScript-style event loop
+/// semantics.
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its state for the current thread.
 pub fn queue_microtask<F>(task: F)
 where
     F: FnOnce() + 'static,
@@ -79,6 +116,11 @@ where
         .push_back(Box::new(task));
 }
 
+/// Schedules a one-shot timer on the current runtime thread.
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its state for the current thread.
 pub fn set_timeout<F>(delay: Duration, callback: F) -> TimeoutHandle
 where
     F: FnOnce() + 'static,
@@ -98,10 +140,22 @@ where
     }
 }
 
+/// Cancels a timeout previously created by [`set_timeout`].
+///
+/// # Panics
+///
+/// Panics if called from a different runtime thread than the one that created `handle`.
 pub fn clear_timeout(handle: &TimeoutHandle) {
     clear_timer(handle.owner, handle.id);
 }
 
+/// Schedules a repeating timer on the current runtime thread.
+///
+/// The callback is invoked once per interval until the handle is cleared.
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its state for the current thread.
 pub fn set_interval<F>(delay: Duration, callback: F) -> IntervalHandle
 where
     F: FnMut() + 'static,
@@ -126,10 +180,33 @@ where
     }
 }
 
+/// Cancels an interval previously created by [`set_interval`].
+///
+/// # Panics
+///
+/// Panics if called from a different runtime thread than the one that created `handle`.
 pub fn clear_interval(handle: &IntervalHandle) {
     clear_timer(handle.owner, handle.id);
 }
 
+/// Queues a future on the current runtime thread.
+///
+/// The future is scheduled immediately and can be awaited through the returned [`JoinHandle`].
+///
+/// The future will be driven to completion regardless or whether the join handle is polled or dropped, so this function
+/// can be used as a convenient way to spawn detached async tasks on the current thread.
+///
+/// # Examples
+///
+/// ```
+/// # let _ = || {
+/// let handle = ruin_runtime::queue_future(async { 42usize });
+/// # };
+/// ```
+///
+/// # Panics
+///
+/// Panics if the runtime cannot initialize its state for the current thread.
 pub fn queue_future<F>(future: F) -> JoinHandle<F::Output>
 where
     F: Future + 'static,
@@ -150,13 +227,21 @@ where
     JoinHandle { state }
 }
 
+/// Spawns a worker runtime thread.
+///
+/// `initial_task` is queued onto the worker as its first macrotask. `on_exit` runs on the parent
+/// runtime thread after the worker shuts down.
+///
+/// # Panics
+///
+/// Panics if the worker thread or its driver cannot be created.
 pub fn spawn_worker<Init, Exit>(initial_task: Init, on_exit: Exit) -> WorkerHandle
 where
     Init: FnOnce() + Send + 'static,
     Exit: FnOnce() + 'static,
 {
     let parent = current_thread();
-    let (reactor, notifier) = create_reactor().expect("worker reactor should initialize");
+    let (driver, notifier) = create_driver().expect("worker driver should initialize");
     let shared = Arc::new(ThreadShared::new(notifier));
     let handle = ThreadHandle {
         shared: Arc::clone(&shared),
@@ -175,7 +260,7 @@ where
     std::thread::Builder::new()
         .name("ruin-runtime-worker".into())
         .spawn(move || {
-            install_thread(shared, reactor, Some(worker_completion));
+            install_thread(shared, driver, Some(worker_completion));
             queue_task(initial_task);
             run();
         })
@@ -187,6 +272,14 @@ where
     }
 }
 
+/// Runs the current runtime thread until no work, timers, child workers, or async operations
+/// remain.
+///
+/// This is the main event loop entry point used by the proc-macro entry attributes.
+///
+/// # Panics
+///
+/// Panics if runtime initialization fails or if the underlying driver returns an unexpected error.
 pub fn run() {
     let _ = current_thread();
 
@@ -223,7 +316,7 @@ pub fn run() {
 
         if has_pending_timers() || state.has_live_children() || state.has_live_async_operations() {
             state.shared.closing.store(false, Ordering::Release);
-            state.reactor.wait().expect("reactor wait should succeed");
+            state.driver.wait().expect("driver wait should succeed");
             continue;
         }
 
@@ -240,16 +333,20 @@ pub fn run() {
 }
 
 fn drain_all() {
-    drain_reactor_events();
+    drain_driver_events();
     drain_remote_tasks();
     drain_completed_workers();
 }
 
+/// Returns a future that yields back to the runtime scheduler once.
 pub fn yield_now() -> YieldNow {
     YieldNow { yielded: false }
 }
 
 impl ThreadHandle {
+    /// Queues a macrotask onto this runtime thread.
+    ///
+    /// Returns `false` if the target thread is already closed.
     pub fn queue_task<F>(&self, task: F) -> bool
     where
         F: FnOnce() + Send + 'static,
@@ -257,6 +354,9 @@ impl ThreadHandle {
         self.shared.enqueue_macro(Box::new(task))
     }
 
+    /// Queues a microtask onto this runtime thread.
+    ///
+    /// Returns `false` if the target thread is already closed.
     pub fn queue_microtask<F>(&self, task: F) -> bool
     where
         F: FnOnce() + Send + 'static,
@@ -264,6 +364,7 @@ impl ThreadHandle {
         self.shared.enqueue_micro(Box::new(task))
     }
 
+    /// Returns `true` if the target runtime thread has shut down.
     pub fn is_closed(&self) -> bool {
         self.shared.closed.load(Ordering::Acquire)
     }
@@ -282,6 +383,9 @@ impl ThreadHandle {
 }
 
 impl WorkerHandle {
+    /// Queues a macrotask onto the worker thread.
+    ///
+    /// Returns `false` if the worker has already shut down.
     pub fn queue_task<F>(&self, task: F) -> bool
     where
         F: FnOnce() + Send + 'static,
@@ -289,6 +393,9 @@ impl WorkerHandle {
         self.thread.queue_task(task)
     }
 
+    /// Queues a microtask onto the worker thread.
+    ///
+    /// Returns `false` if the worker has already shut down.
     pub fn queue_microtask<F>(&self, task: F) -> bool
     where
         F: FnOnce() + Send + 'static,
@@ -296,10 +403,12 @@ impl WorkerHandle {
         self.thread.queue_microtask(task)
     }
 
+    /// Returns `true` once the worker thread has fully exited.
     pub fn is_finished(&self) -> bool {
         self.completion.finished.load(Ordering::Acquire)
     }
 
+    /// Returns a generic [`ThreadHandle`] for the worker thread.
     pub fn thread(&self) -> ThreadHandle {
         self.thread.clone()
     }
@@ -328,7 +437,7 @@ impl Future for YieldNow {
 }
 
 struct ThreadState {
-    reactor: Reactor,
+    driver: Driver,
     shared: Arc<ThreadShared>,
     worker_completion: Option<Arc<WorkerCompletion>>,
     local_microtasks: RefCell<VecDeque<LocalTask>>,
@@ -341,11 +450,11 @@ struct ThreadState {
 impl ThreadState {
     fn new(
         shared: Arc<ThreadShared>,
-        reactor: Reactor,
+        driver: Driver,
         worker_completion: Option<Arc<WorkerCompletion>>,
     ) -> Self {
         Self {
-            reactor,
+            driver,
             shared,
             worker_completion,
             local_microtasks: RefCell::new(VecDeque::new()),
@@ -695,11 +804,11 @@ unsafe fn future_task_drop(data: *const ()) {
 fn current_thread() -> &'static ThreadState {
     unsafe {
         if CURRENT_THREAD.is_null() {
-            let (reactor, notifier) = create_reactor().expect("runtime reactor should initialize");
+            let (driver, notifier) = create_driver().expect("runtime driver should initialize");
             let shared = Arc::new(ThreadShared::new(notifier));
-            let state = Box::new(ThreadState::new(shared, reactor, None));
+            let state = Box::new(ThreadState::new(shared, driver, None));
             let state = Box::into_raw(state);
-            (*state).reactor.bind_current_thread();
+            (*state).driver.bind_current_thread();
             CURRENT_THREAD = state;
         }
 
@@ -713,14 +822,14 @@ fn current_thread_ptr() -> *const ThreadState {
 
 fn install_thread(
     shared: Arc<ThreadShared>,
-    reactor: Reactor,
+    driver: Driver,
     worker_completion: Option<Arc<WorkerCompletion>>,
 ) {
     unsafe {
         debug_assert!(CURRENT_THREAD.is_null(), "thread runtime already installed");
-        let state = Box::new(ThreadState::new(shared, reactor, worker_completion));
+        let state = Box::new(ThreadState::new(shared, driver, worker_completion));
         let state = Box::into_raw(state);
-        (*state).reactor.bind_current_thread();
+        (*state).driver.bind_current_thread();
         CURRENT_THREAD = state;
     }
 }
@@ -731,18 +840,18 @@ fn teardown_thread() {
         CURRENT_THREAD = ptr::null_mut();
 
         if !state.is_null() {
-            (*state).reactor.unbind_current_thread();
+            (*state).driver.unbind_current_thread();
             drop(Box::from_raw(state));
         }
     }
 }
 
-fn drain_reactor_events() {
+fn drain_driver_events() {
     loop {
         let ready = current_thread()
-            .reactor
+            .driver
             .poll()
-            .expect("reactor poll should succeed");
+            .expect("driver poll should succeed");
 
         let Some(ready) = ready else {
             break;
@@ -751,13 +860,13 @@ fn drain_reactor_events() {
         let state = current_thread();
         if ready.wake {
             let _ = state
-                .reactor
+                .driver
                 .drain_wake()
                 .expect("wake drain should succeed");
         }
         if ready.timer {
             let _ = state
-                .reactor
+                .driver
                 .drain_timer()
                 .expect("timer drain should succeed");
             dispatch_expired_timers();
@@ -906,9 +1015,9 @@ fn dispatch_expired_timers() {
 fn rearm_thread_timer() {
     let deadline = current_thread().timers.borrow().peek_deadline();
     current_thread()
-        .reactor
+        .driver
         .rearm_timer(deadline)
-        .expect("timerfd rearm should succeed");
+        .expect("driver timer rearm should succeed");
 }
 
 fn deadline_from_now(delay: Duration) -> Duration {
diff --git a/lib/runtime/src/sys/linux/fs.rs b/lib/runtime/src/sys/linux/fs.rs
index 84e39da..88aef7c 100644
--- a/lib/runtime/src/sys/linux/fs.rs
+++ b/lib/runtime/src/sys/linux/fs.rs
@@ -16,7 +16,7 @@ use std::thread;
 use crate::op::completion::completion_for_current_thread;
 use crate::op::fs::{FileType, FsOp, MetadataTarget, OpenOptions, RawDirEntry, RawMetadata};
 use crate::platform::linux_x86_64::runtime::{
-    ThreadHandle, current_thread_handle, with_current_reactor,
+    ThreadHandle, current_thread_handle, with_current_driver,
 };
 use crate::platform::linux_x86_64::uring::{
     IORING_FSYNC_DATASYNC, IORING_OP_CLOSE, IORING_OP_FSYNC, IORING_OP_FTRUNCATE,
@@ -474,14 +474,14 @@ where
 {
     let (future, handle) = completion_for_current_thread::<io::Result<T>>();
     let callback_handle = handle.clone();
-    let token = with_current_reactor(|reactor| {
-        reactor.submit_operation(fill, move |cqe| {
+    let token = with_current_driver(|driver| {
+        driver.submit_operation(fill, move |cqe| {
             callback_handle.complete(map(cqe));
         })
     })?;
 
     handle.set_cancel(move || {
-        let _ = with_current_reactor(|reactor| reactor.cancel_operation(token));
+        let _ = with_current_driver(|driver| driver.cancel_operation(token));
     });
 
     future.await
diff --git a/lib/runtime/src/sys/linux/net.rs b/lib/runtime/src/sys/linux/net.rs
index 85111dc..df81b87 100644
--- a/lib/runtime/src/sys/linux/net.rs
+++ b/lib/runtime/src/sys/linux/net.rs
@@ -14,7 +14,7 @@ use std::time::Duration;
 
 use crate::op::completion::completion_for_current_thread;
 use crate::op::net::{AcceptedSocket, NetOp, ReceivedDatagram};
-use crate::platform::linux_x86_64::runtime::with_current_reactor;
+use crate::platform::linux_x86_64::runtime::with_current_driver;
 use crate::platform::linux_x86_64::uring::{
     IORING_OP_ACCEPT, IORING_OP_BIND, IORING_OP_CLOSE, IORING_OP_CONNECT, IORING_OP_LISTEN,
     IORING_OP_RECV, IORING_OP_SEND, IORING_OP_SHUTDOWN, IORING_OP_SOCKET, IoUringCqe, IoUringSqe,
@@ -556,14 +556,14 @@ where
 {
     let (future, handle) = completion_for_current_thread::<io::Result<T>>();
     let callback_handle = handle.clone();
-    let token = with_current_reactor(|reactor| {
-        reactor.submit_operation(fill, move |cqe| {
+    let token = with_current_driver(|driver| {
+        driver.submit_operation(fill, move |cqe| {
             callback_handle.complete(map(cqe));
         })
     })?;
 
     handle.set_cancel(move || {
-        let _ = with_current_reactor(|reactor| reactor.cancel_operation(token));
+        let _ = with_current_driver(|driver| driver.cancel_operation(token));
     });
 
     future.await
diff --git a/lib/runtime/src/time.rs b/lib/runtime/src/time.rs
index 35cda8c..930611c 100644
--- a/lib/runtime/src/time.rs
+++ b/lib/runtime/src/time.rs
@@ -1,9 +1,11 @@
 //! Runtime time primitives.
+//!
+//! These helpers integrate with the runtime's timer queue and are designed to be used from
+//! futures scheduled with [`crate::queue_future`] or one of the runtime entry macros.
 
 use std::cell::{Cell, RefCell};
 use std::fmt;
 use std::future::{Future, poll_fn};
-use std::io;
 use std::pin::Pin;
 use std::rc::Rc;
 use std::task::Waker;
@@ -12,6 +14,7 @@ use std::time::Duration;
 
 use crate::{clear_timeout, set_timeout};
 
+/// Future returned by [`sleep`].
 pub struct Sleep {
     delay: Option<Duration>,
     state: Option<Rc<SleepState>>,
@@ -20,8 +23,18 @@ pub struct Sleep {
 }
 
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
+/// Error returned by [`timeout`] when the deadline expires first.
 pub struct Elapsed;
 
+/// Returns a future that completes after `duration` has elapsed on the current runtime thread.
+///
+/// # Examples
+///
+/// ```
+/// # let _ = || async {
+/// ruin_runtime::time::sleep(std::time::Duration::from_millis(10)).await;
+/// # };
+/// ```
 pub fn sleep(duration: Duration) -> Sleep {
     Sleep {
         delay: Some(duration),
@@ -31,6 +44,24 @@ pub fn sleep(duration: Duration) -> Sleep {
     }
 }
 
+/// Runs `future` until it completes or `duration` elapses, whichever happens first.
+///
+/// The wrapped future is dropped when the timeout fires. As with other runtime operations, dropping
+/// a future cancels interest in the result but does not guarantee cancellation of any underlying
+/// OS work that future may have started.
+///
+/// # Examples
+///
+/// ```
+/// # let _ = || async {
+/// let result = ruin_runtime::time::timeout(
+///     std::time::Duration::from_millis(5),
+///     async { 42usize },
+/// )
+/// .await;
+/// assert_eq!(result, Ok(42));
+/// # };
+/// ```
 pub async fn timeout<F>(duration: Duration, future: F) -> Result<F::Output, Elapsed>
 where
     F: Future,
@@ -52,10 +83,6 @@ where
     .await
 }
 
-pub fn timeout_error(action: &'static str) -> io::Error {
-    io::Error::new(io::ErrorKind::TimedOut, format!("{action} timed out"))
-}
-
 impl Future for Sleep {
     type Output = ();