core: add a sandbox-backed fs helper

This commit is contained in:
Michael Bolin
2026-03-19 09:45:52 -07:00
parent 859c58f07d
commit 2e1d27522b
18 changed files with 674 additions and 0 deletions

13
codex-rs/Cargo.lock generated
View File

@@ -1557,6 +1557,7 @@ version = "0.0.0"
dependencies = [
"anyhow",
"codex-apply-patch",
"codex-fs-ops",
"codex-linux-sandbox",
"codex-shell-escalation",
"codex-utils-home-dir",
@@ -1846,6 +1847,7 @@ dependencies = [
"codex-exec-server",
"codex-execpolicy",
"codex-file-search",
"codex-fs-ops",
"codex-git",
"codex-hooks",
"codex-keyring-store",
@@ -2087,6 +2089,17 @@ dependencies = [
"tokio",
]
[[package]]
name = "codex-fs-ops"
version = "0.0.0"
dependencies = [
"anyhow",
"pretty_assertions",
"serde",
"serde_json",
"tempfile",
]
[[package]]
name = "codex-git"
version = "0.0.0"

View File

@@ -11,6 +11,7 @@ members = [
"apply-patch",
"arg0",
"feedback",
"fs-ops",
"codex-backend-openapi-models",
"cloud-requirements",
"cloud-tasks",
@@ -109,6 +110,7 @@ codex-exec-server = { path = "exec-server" }
codex-execpolicy = { path = "execpolicy" }
codex-experimental-api-macros = { path = "codex-experimental-api-macros" }
codex-feedback = { path = "feedback" }
codex-fs-ops = { path = "fs-ops" }
codex-file-search = { path = "file-search" }
codex-git = { path = "utils/git" }
codex-hooks = { path = "hooks" }

View File

@@ -14,6 +14,7 @@ workspace = true
[dependencies]
anyhow = { workspace = true }
codex-apply-patch = { workspace = true }
codex-fs-ops = { workspace = true }
codex-linux-sandbox = { workspace = true }
codex-shell-escalation = { workspace = true }
codex-utils-home-dir = { workspace = true }

View File

@@ -4,6 +4,7 @@ use std::path::Path;
use std::path::PathBuf;
use codex_apply_patch::CODEX_CORE_APPLY_PATCH_ARG1;
use codex_fs_ops::CODEX_CORE_FS_OPS_ARG1;
use codex_utils_home_dir::find_codex_home;
#[cfg(unix)]
use std::os::unix::fs::symlink;
@@ -105,6 +106,17 @@ pub fn arg0_dispatch() -> Option<Arg0PathEntryGuard> {
};
std::process::exit(exit_code);
}
if argv1 == CODEX_CORE_FS_OPS_ARG1 {
let mut stdin = std::io::stdin();
let mut stdout = std::io::stdout();
let mut stderr = std::io::stderr();
let exit_code =
match codex_fs_ops::run_from_args(args, &mut stdin, &mut stdout, &mut stderr) {
Ok(()) => 0,
Err(_) => 1,
};
std::process::exit(exit_code);
}
// This modifies the environment, which is not thread-safe, so do this
// before creating any threads/the Tokio runtime.

View File

@@ -38,6 +38,7 @@ codex-exec-server = { workspace = true }
codex-shell-command = { workspace = true }
codex-skills = { workspace = true }
codex-execpolicy = { workspace = true }
codex-fs-ops = { workspace = true }
codex-file-search = { workspace = true }
codex-git = { workspace = true }
codex-hooks = { workspace = true }

View File

@@ -332,6 +332,58 @@ pub(crate) async fn execute_exec_request(
finalize_exec_result(raw_output_result, sandbox, duration)
}
pub(crate) async fn execute_exec_request_output_bytes(
exec_request: ExecRequest,
sandbox_policy: &SandboxPolicy,
stdout_stream: Option<StdoutStream>,
after_spawn: Option<Box<dyn FnOnce() + Send>>,
) -> Result<ExecToolCallOutputBytes> {
let ExecRequest {
command,
cwd,
env,
network,
expiration,
sandbox,
windows_sandbox_level,
windows_sandbox_private_desktop,
sandbox_permissions,
sandbox_policy: _sandbox_policy_from_env,
file_system_sandbox_policy,
network_sandbox_policy,
justification,
arg0,
} = exec_request;
let _ = _sandbox_policy_from_env;
let params = ExecParams {
command,
cwd,
expiration,
env,
network: network.clone(),
sandbox_permissions,
windows_sandbox_level,
windows_sandbox_private_desktop,
justification,
arg0,
};
let start = Instant::now();
let raw_output_result = exec(
params,
sandbox,
sandbox_policy,
&file_system_sandbox_policy,
network_sandbox_policy,
stdout_stream,
after_spawn,
)
.await;
let duration = start.elapsed();
finalize_exec_result_bytes(raw_output_result, sandbox, duration)
}
#[cfg(target_os = "windows")]
fn extract_create_process_as_user_error_code(err: &str) -> Option<String> {
let marker = "CreateProcessAsUserW failed: ";
@@ -574,6 +626,64 @@ fn finalize_exec_result(
}
}
fn finalize_exec_result_bytes(
raw_output_result: std::result::Result<RawExecToolCallOutput, CodexErr>,
sandbox_type: SandboxType,
duration: Duration,
) -> Result<ExecToolCallOutputBytes> {
match raw_output_result {
Ok(raw_output) => {
#[allow(unused_mut)]
let mut timed_out = raw_output.timed_out;
#[cfg(target_family = "unix")]
{
if let Some(signal) = raw_output.exit_status.signal() {
if signal == TIMEOUT_CODE {
timed_out = true;
} else {
return Err(CodexErr::Sandbox(SandboxErr::Signal(signal)));
}
}
}
let mut exit_code = raw_output.exit_status.code().unwrap_or(-1);
if timed_out {
exit_code = EXEC_TIMEOUT_EXIT_CODE;
}
let exec_output = ExecToolCallOutputBytes {
exit_code,
stdout: raw_output.stdout,
stderr: raw_output.stderr,
aggregated_output: raw_output.aggregated_output,
duration,
timed_out,
};
if timed_out {
return Err(CodexErr::Sandbox(SandboxErr::Timeout {
output: Box::new(exec_output.to_utf8_lossy_output()),
}));
}
let string_output = exec_output.to_utf8_lossy_output();
if is_likely_sandbox_denied(sandbox_type, &string_output) {
return Err(CodexErr::Sandbox(SandboxErr::Denied {
output: Box::new(string_output),
network_policy_decision: None,
}));
}
Ok(exec_output)
}
Err(err) => {
tracing::error!("exec error: {err}");
Err(err)
}
}
}
pub(crate) mod errors {
use super::CodexErr;
use crate::sandboxing::SandboxTransformError;
@@ -741,6 +851,16 @@ pub struct ExecToolCallOutput {
pub timed_out: bool,
}
#[derive(Clone, Debug)]
pub(crate) struct ExecToolCallOutputBytes {
pub exit_code: i32,
pub stdout: StreamOutput<Vec<u8>>,
pub stderr: StreamOutput<Vec<u8>>,
pub aggregated_output: StreamOutput<Vec<u8>>,
pub duration: Duration,
pub timed_out: bool,
}
impl Default for ExecToolCallOutput {
fn default() -> Self {
Self {
@@ -754,6 +874,19 @@ impl Default for ExecToolCallOutput {
}
}
impl ExecToolCallOutputBytes {
fn to_utf8_lossy_output(&self) -> ExecToolCallOutput {
ExecToolCallOutput {
exit_code: self.exit_code,
stdout: self.stdout.from_utf8_lossy(),
stderr: self.stderr.from_utf8_lossy(),
aggregated_output: self.aggregated_output.from_utf8_lossy(),
duration: self.duration,
timed_out: self.timed_out,
}
}
}
#[cfg_attr(not(target_os = "windows"), allow(unused_variables))]
async fn exec(
params: ExecParams,

View File

@@ -114,6 +114,7 @@ pub mod default_client;
pub mod project_doc;
mod rollout;
pub(crate) mod safety;
mod sandboxed_fs;
pub mod seatbelt;
pub mod shell;
pub mod shell_snapshot;

View File

@@ -0,0 +1,197 @@
use crate::codex::Session;
use crate::codex::TurnContext;
use crate::error::CodexErr;
use crate::error::SandboxErr;
use crate::exec::ExecExpiration;
use crate::exec::ExecToolCallOutputBytes;
use crate::sandboxing::CommandSpec;
use crate::sandboxing::SandboxPermissions;
use crate::sandboxing::execute_env_output_bytes;
use crate::sandboxing::merge_permission_profiles;
use crate::tools::sandboxing::SandboxAttempt;
use crate::tools::sandboxing::SandboxablePreference;
use codex_fs_ops::CODEX_CORE_FS_OPS_ARG1;
use codex_fs_ops::FsError;
use codex_fs_ops::FsErrorKind;
use codex_protocol::models::PermissionProfile;
use std::collections::HashMap;
use std::path::Path;
use std::path::PathBuf;
use std::sync::Arc;
use std::time::Duration;
const SANDBOXED_FS_TIMEOUT: Duration = Duration::from_secs(30);
pub(crate) async fn read_file(
session: &Arc<Session>,
turn: &Arc<TurnContext>,
path: &Path,
) -> Result<Vec<u8>, SandboxedFsError> {
let output = run_request(session, turn, path).await?;
Ok(output.stdout.text)
}
async fn run_request(
session: &Arc<Session>,
turn: &Arc<TurnContext>,
path: &Path,
) -> Result<ExecToolCallOutputBytes, SandboxedFsError> {
let exe = std::env::current_exe().map_err(|error| SandboxedFsError::ResolveExe {
message: error.to_string(),
})?;
let additional_permissions = effective_granted_permissions(session).await;
let sandbox_manager = crate::sandboxing::SandboxManager::new();
let attempt = SandboxAttempt {
sandbox: sandbox_manager.select_initial(
&turn.file_system_sandbox_policy,
turn.network_sandbox_policy,
SandboxablePreference::Auto,
turn.windows_sandbox_level,
/*has_managed_network_requirements*/ false,
),
policy: &turn.sandbox_policy,
file_system_policy: &turn.file_system_sandbox_policy,
network_policy: turn.network_sandbox_policy,
enforce_managed_network: false,
manager: &sandbox_manager,
sandbox_cwd: &turn.cwd,
codex_linux_sandbox_exe: turn.codex_linux_sandbox_exe.as_ref(),
use_legacy_landlock: turn.features.use_legacy_landlock(),
windows_sandbox_level: turn.windows_sandbox_level,
windows_sandbox_private_desktop: turn.config.permissions.windows_sandbox_private_desktop,
};
let exec_request = attempt
.env_for(
CommandSpec {
program: exe.to_string_lossy().to_string(),
args: vec![
CODEX_CORE_FS_OPS_ARG1.to_string(),
"read".to_string(),
path.to_string_lossy().to_string(),
],
cwd: turn.cwd.clone(),
env: HashMap::new(),
expiration: ExecExpiration::Timeout(SANDBOXED_FS_TIMEOUT),
sandbox_permissions: SandboxPermissions::UseDefault,
additional_permissions,
justification: None,
},
/*network*/ None,
)
.map_err(|error| SandboxedFsError::ProcessFailed {
path: path.to_path_buf(),
exit_code: -1,
message: error.to_string(),
})?;
let output = execute_env_output_bytes(exec_request, /*stdout_stream*/ None)
.await
.map_err(|error| map_exec_error(path, error))?;
if output.exit_code != 0 {
return Err(parse_helper_failure(
path,
output.exit_code,
&output.stderr.text,
&output.stdout.text,
));
}
Ok(output)
}
async fn effective_granted_permissions(session: &Session) -> Option<PermissionProfile> {
let granted_session_permissions = session.granted_session_permissions().await;
let granted_turn_permissions = session.granted_turn_permissions().await;
merge_permission_profiles(
granted_session_permissions.as_ref(),
granted_turn_permissions.as_ref(),
)
}
fn map_exec_error(path: &Path, error: CodexErr) -> SandboxedFsError {
match error {
CodexErr::Sandbox(SandboxErr::Timeout { .. }) => SandboxedFsError::TimedOut {
path: path.to_path_buf(),
},
_ => SandboxedFsError::ProcessFailed {
path: path.to_path_buf(),
exit_code: -1,
message: error.to_string(),
},
}
}
fn parse_helper_failure(
path: &Path,
exit_code: i32,
stderr: &[u8],
stdout: &[u8],
) -> SandboxedFsError {
if let Ok(error) = serde_json::from_slice::<FsError>(stderr) {
return SandboxedFsError::Operation {
path: path.to_path_buf(),
error,
};
}
let stderr = String::from_utf8_lossy(stderr);
let stdout = String::from_utf8_lossy(stdout);
let message = if !stderr.trim().is_empty() {
stderr.trim().to_string()
} else if !stdout.trim().is_empty() {
stdout.trim().to_string()
} else {
"no error details emitted".to_string()
};
SandboxedFsError::ProcessFailed {
path: path.to_path_buf(),
exit_code,
message,
}
}
#[derive(Debug, thiserror::Error)]
pub(crate) enum SandboxedFsError {
#[error("failed to determine codex executable: {message}")]
ResolveExe { message: String },
#[error("sandboxed fs helper timed out while accessing `{path}`")]
TimedOut { path: PathBuf },
#[error("sandboxed fs helper exited with code {exit_code} while accessing `{path}`: {message}")]
ProcessFailed {
path: PathBuf,
exit_code: i32,
message: String,
},
#[error("sandboxed fs helper could not access `{path}`: {error}")]
Operation { path: PathBuf, error: FsError },
}
impl SandboxedFsError {
pub(crate) fn operation_error_kind(&self) -> Option<&FsErrorKind> {
match self {
Self::Operation { error, .. } => Some(&error.kind),
_ => None,
}
}
pub(crate) fn operation_error_message(&self) -> Option<&str> {
match self {
Self::Operation { error, .. } => Some(error.message.as_str()),
_ => None,
}
}
#[allow(dead_code)]
pub(crate) fn to_io_error(&self) -> std::io::Error {
match self {
Self::Operation { error, .. } => error.to_io_error(),
Self::TimedOut { .. } => {
std::io::Error::new(std::io::ErrorKind::TimedOut, self.to_string())
}
Self::ResolveExe { .. } | Self::ProcessFailed { .. } => {
std::io::Error::other(self.to_string())
}
}
}
}

View File

@@ -10,9 +10,11 @@ pub(crate) mod macos_permissions;
use crate::exec::ExecExpiration;
use crate::exec::ExecToolCallOutput;
use crate::exec::ExecToolCallOutputBytes;
use crate::exec::SandboxType;
use crate::exec::StdoutStream;
use crate::exec::execute_exec_request;
use crate::exec::execute_exec_request_output_bytes;
use crate::landlock::allow_network_for_proxy;
use crate::landlock::create_linux_sandbox_command_args_for_policies;
use crate::protocol::SandboxPolicy;
@@ -738,6 +740,20 @@ pub async fn execute_env(
.await
}
pub(crate) async fn execute_env_output_bytes(
exec_request: ExecRequest,
stdout_stream: Option<StdoutStream>,
) -> crate::error::Result<ExecToolCallOutputBytes> {
let effective_policy = exec_request.sandbox_policy.clone();
execute_exec_request_output_bytes(
exec_request,
&effective_policy,
stdout_stream,
/*after_spawn*/ None,
)
.await
}
pub async fn execute_exec_request_with_after_spawn(
exec_request: ExecRequest,
stdout_stream: Option<StdoutStream>,

View File

@@ -0,0 +1,6 @@
load("//:defs.bzl", "codex_rust_crate")
codex_rust_crate(
name = "fs-ops",
crate_name = "codex_fs_ops",
)

View File

@@ -0,0 +1,21 @@
[package]
name = "codex-fs-ops"
version.workspace = true
edition.workspace = true
license.workspace = true
[lib]
name = "codex_fs_ops"
path = "src/lib.rs"
[lints]
workspace = true
[dependencies]
anyhow = { workspace = true }
serde = { workspace = true, features = ["derive"] }
serde_json = { workspace = true }
[dev-dependencies]
pretty_assertions = { workspace = true }
tempfile = { workspace = true }

View File

@@ -0,0 +1,38 @@
use std::ffi::OsString;
use std::path::PathBuf;
#[derive(Debug, Clone, PartialEq, Eq)]
pub enum FsCommand {
ReadFile { path: PathBuf },
}
pub fn parse_command_from_args(
mut args: impl Iterator<Item = OsString>,
) -> Result<FsCommand, String> {
let Some(operation) = args.next() else {
return Err("missing operation".to_string());
};
let Some(operation) = operation.to_str() else {
return Err("operation must be valid UTF-8".to_string());
};
let Some(path) = args.next() else {
return Err(format!("missing path for operation `{operation}`"));
};
if args.next().is_some() {
return Err(format!(
"unexpected extra arguments for operation `{operation}`"
));
}
let path = PathBuf::from(path);
match operation {
"read" => Ok(FsCommand::ReadFile { path }),
_ => Err(format!(
"unsupported filesystem operation `{operation}`; expected `read`"
)),
}
}
#[cfg(test)]
#[path = "command_tests.rs"]
mod tests;

View File

@@ -0,0 +1,16 @@
use super::FsCommand;
use super::parse_command_from_args;
use pretty_assertions::assert_eq;
#[test]
fn parse_read_command() {
let command = parse_command_from_args(["read", "/tmp/example.png"].into_iter().map(Into::into))
.expect("command should parse");
assert_eq!(
command,
FsCommand::ReadFile {
path: "/tmp/example.png".into(),
}
);
}

View File

@@ -0,0 +1,3 @@
/// Special argv[1] flag used when the Codex executable self-invokes to run the
/// internal sandbox-backed filesystem helper path.
pub const CODEX_CORE_FS_OPS_ARG1: &str = "--codex-run-as-fs-ops";

View File

@@ -0,0 +1,70 @@
use serde::Deserialize;
use serde::Serialize;
use std::io::ErrorKind;
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "snake_case")]
pub enum FsErrorKind {
NotFound,
PermissionDenied,
IsADirectory,
InvalidData,
Other,
}
impl From<ErrorKind> for FsErrorKind {
fn from(value: ErrorKind) -> Self {
match value {
ErrorKind::NotFound => Self::NotFound,
ErrorKind::PermissionDenied => Self::PermissionDenied,
ErrorKind::IsADirectory => Self::IsADirectory,
ErrorKind::InvalidData => Self::InvalidData,
_ => Self::Other,
}
}
}
impl FsErrorKind {
pub fn to_io_error_kind(&self) -> ErrorKind {
match self {
Self::NotFound => ErrorKind::NotFound,
Self::PermissionDenied => ErrorKind::PermissionDenied,
Self::IsADirectory => ErrorKind::IsADirectory,
Self::InvalidData => ErrorKind::InvalidData,
Self::Other => ErrorKind::Other,
}
}
}
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
pub struct FsError {
pub kind: FsErrorKind,
pub message: String,
pub raw_os_error: Option<i32>,
}
impl std::fmt::Display for FsError {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.write_str(&self.message)
}
}
impl FsError {
pub fn to_io_error(&self) -> std::io::Error {
if let Some(raw_os_error) = self.raw_os_error {
std::io::Error::from_raw_os_error(raw_os_error)
} else {
std::io::Error::new(self.kind.to_io_error_kind(), self.message.clone())
}
}
}
impl From<std::io::Error> for FsError {
fn from(error: std::io::Error) -> Self {
Self {
kind: error.kind().into(),
message: error.to_string(),
raw_os_error: error.raw_os_error(),
}
}
}

View File

@@ -0,0 +1,13 @@
mod command;
mod constants;
mod error;
mod runner;
pub use command::FsCommand;
pub use command::parse_command_from_args;
pub use constants::CODEX_CORE_FS_OPS_ARG1;
pub use error::FsError;
pub use error::FsErrorKind;
pub use runner::execute;
pub use runner::run_from_args;
pub use runner::write_error;

View File

@@ -0,0 +1,55 @@
use crate::FsCommand;
use crate::FsError;
use crate::parse_command_from_args;
use anyhow::Context;
use anyhow::Result;
use std::ffi::OsString;
use std::io::Read;
use std::io::Write;
pub fn run_from_args(
args: impl Iterator<Item = OsString>,
stdin: &mut impl Read,
stdout: &mut impl Write,
stderr: &mut impl Write,
) -> Result<()> {
let command = match parse_command_from_args(args) {
Ok(command) => command,
Err(error) => {
writeln!(stderr, "{error}").context("failed to write fs helper usage error")?;
anyhow::bail!("{error}");
}
};
if let Err(error) = execute(command, stdin, stdout) {
write_error(stderr, &error)?;
anyhow::bail!("{error}");
}
Ok(())
}
pub fn execute(
command: FsCommand,
_stdin: &mut impl Read,
stdout: &mut impl Write,
) -> Result<(), FsError> {
match command {
FsCommand::ReadFile { path } => {
let mut file = std::fs::File::open(path).map_err(FsError::from)?;
std::io::copy(&mut file, stdout)
.map(|_| ())
.map_err(FsError::from)
}
}
}
pub fn write_error(stderr: &mut impl Write, error: &FsError) -> Result<()> {
serde_json::to_writer(&mut *stderr, error).context("failed to serialize fs error")?;
writeln!(stderr).context("failed to terminate fs error with newline")?;
Ok(())
}
#[cfg(test)]
#[path = "runner_tests.rs"]
mod tests;

View File

@@ -0,0 +1,76 @@
use super::execute;
use crate::FsCommand;
use crate::FsErrorKind;
use crate::run_from_args;
use pretty_assertions::assert_eq;
use tempfile::tempdir;
#[test]
fn run_from_args_streams_file_bytes_to_stdout() {
let tempdir = tempdir().expect("tempdir");
let path = tempdir.path().join("image.bin");
let expected = b"hello\x00world".to_vec();
std::fs::write(&path, &expected).expect("write test file");
let mut stdout = Vec::new();
let mut stderr = Vec::new();
let mut stdin = std::io::empty();
run_from_args(
["read", path.to_str().expect("utf-8 test path")]
.into_iter()
.map(Into::into),
&mut stdin,
&mut stdout,
&mut stderr,
)
.expect("read should succeed");
assert_eq!(stdout, expected);
assert_eq!(stderr, Vec::<u8>::new());
}
#[test]
fn read_reports_directory_error() {
let tempdir = tempdir().expect("tempdir");
let mut stdout = Vec::new();
let mut stdin = std::io::empty();
let error = execute(
FsCommand::ReadFile {
path: tempdir.path().to_path_buf(),
},
&mut stdin,
&mut stdout,
)
.expect_err("reading a directory should fail");
#[cfg(target_os = "windows")]
assert_eq!(error.kind, FsErrorKind::PermissionDenied);
#[cfg(not(target_os = "windows"))]
assert_eq!(error.kind, FsErrorKind::IsADirectory);
}
#[test]
fn run_from_args_serializes_errors_to_stderr() {
let tempdir = tempdir().expect("tempdir");
let missing = tempdir.path().join("missing.txt");
let mut stdout = Vec::new();
let mut stderr = Vec::new();
let mut stdin = std::io::empty();
let result = run_from_args(
["read", missing.to_str().expect("utf-8 test path")]
.into_iter()
.map(Into::into),
&mut stdin,
&mut stdout,
&mut stderr,
);
assert!(result.is_err(), "missing file should fail");
assert_eq!(stdout, Vec::<u8>::new());
let error: crate::FsError = serde_json::from_slice(&stderr).expect("structured fs error");
assert_eq!(error.kind, FsErrorKind::NotFound);
assert!(error.raw_os_error.is_some());
}