## Why
`zsh-fork` sessions launched through unified-exec need the escalation
socket to survive the wrapper -> server -> child handoff so later
intercepted `exec()` calls can still reach the escalation server.
The inherited-fd spawn path also needs to avoid closing Rust's internal
exec-error pipe, and the shell-escalation handoff needs to tolerate the
receive-side case where a transferred fd is installed into the same
stdio slot it will be mapped onto.
## What Changed
- Added `SpawnLifecycle::inherited_fds()` in
`codex-rs/core/src/unified_exec/process.rs` and threaded inherited fds
through `codex-rs/core/src/unified_exec/process_manager.rs` so
unified-exec can preserve required descriptors across both PTY and
no-stdin pipe spawn paths.
- Updated `codex-rs/core/src/tools/runtimes/shell/zsh_fork_backend.rs`
to expose the escalation socket fd through the spawn lifecycle.
- Added inherited-fd-aware spawn helpers in
`codex-rs/utils/pty/src/pty.rs` and `codex-rs/utils/pty/src/pipe.rs`,
including Unix pre-exec fd pruning that preserves requested inherited
fds while leaving `FD_CLOEXEC` descriptors alone. The pruning helper is
now named `close_inherited_fds_except()` to better describe that
behavior.
- Updated `codex-rs/shell-escalation/src/unix/escalate_client.rs` to
duplicate local stdio before transfer and send destination stdio numbers
in `SuperExecMessage`, so the wrapper keeps using its own
`stdin`/`stdout`/`stderr` until the escalated child takes over.
- Updated `codex-rs/shell-escalation/src/unix/escalate_server.rs` so the
server accepts the overlap case where a received fd reuses the same
stdio descriptor number that the child setup will target with `dup2`.
- Added comments around the PTY stdio wiring and the overlap regression
helper to make the fd handoff and controlling-terminal setup easier to
follow.
## Verification
- `cargo test -p codex-utils-pty`
- covers preserved-fd PTY spawn behavior, PTY resize, Python REPL
continuity, exec-failure reporting, and the no-stdin pipe path
- `cargo test -p codex-shell-escalation`
- covers duplicated-fd transfer on the client side and verifies the
overlap case by passing a pipe-backed stdin payload through the
server-side `dup2` path
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/13644).
* #14624
* __->__ #13644
(Experimental)
This PR adds a first MVP for hooks, with SessionStart and Stop
The core design is:
- hooks live in a dedicated engine under codex-rs/hooks
- each hook type has its own event-specific file
- hook execution is synchronous and blocks normal turn progression while
running
- matching hooks run in parallel, then their results are aggregated into
a normalized HookRunSummary
On the AppServer side, hooks are exposed as operational metadata rather
than transcript-native items:
- new live notifications: hook/started, hook/completed
- persisted/replayed hook results live on Turn.hookRuns
- we intentionally did not add hook-specific ThreadItem variants
Hooks messages are not persisted, they remain ephemeral. The context
changes they add are (they get appended to the user's prompt)
## Summary
- restore the guardian review request snapshot test and its tracked
snapshot after it was dropped from `main`
- make Bazel Rust unit-test wrappers resolve runfiles correctly on
manifest-only platforms like macOS and point Insta at the real workspace
root
- harden the shell-escalation socket-closure assertion so the musl Bazel
test no longer depends on fd reuse behavior
## Verification
- cargo test -p codex-core
guardian_review_request_layout_matches_model_visible_request_snapshot
- cargo test -p codex-shell-escalation
- bazel test //codex-rs/exec:exec-unit-tests
//codex-rs/shell-escalation:shell-escalation-unit-tests
Supersedes #13894.
---------
Co-authored-by: Ahmed Ibrahim <aibrahim@openai.com>
Co-authored-by: viyatb-oai <viyatb@openai.com>
Co-authored-by: Codex <noreply@openai.com>
## Why
`shell_zsh_fork` already provides stronger guarantees around which
executables receive elevated permissions. To reuse that machinery from
unified exec without pushing Unix-specific escalation details through
generic runtime code, the escalation bootstrap and session lifetime
handling need a cleaner boundary.
That boundary also needs to be safe for long-lived sessions: when an
intercepted shell session is closed or pruned, any in-flight approval
workers and any already-approved escalated child they spawned must be
torn down with the session, and the inherited escalation socket must not
leak into unrelated subprocesses.
## What Changed
- Extracted a reusable `EscalationSession` and
`EscalateServer::start_session(...)` in `shell-escalation` so callers
can get the wrapper/socket env overlay and keep the escalation server
alive without immediately running a one-shot command.
- Documented that `EscalationSession::env()` and
`ShellCommandExecutor::run(...)` exchange only that env overlay, which
callers must merge into their own base shell environment.
- Clarified the prepared-exec helper boundary in `core` by naming the
new helper APIs around `ExecRequest`, while keeping the legacy
`execute_env(...)` entrypoints as thin compatibility wrappers for
existing callers that still use the older naming.
- Added a small post-spawn hook on the prepared execution path so the
parent copy of the inheritable escalation socket is closed immediately
after both the existing one-shot shell-command spawn and the
unified-exec spawn.
- Made session teardown explicit with session-scoped cancellation:
dropping an `EscalationSession` or canceling its parent request now
stops intercept workers, and the server-spawned escalated child uses
`kill_on_drop(true)` so teardown cannot orphan an already-approved
child.
- Added `UnifiedExecBackendConfig` plumbing through `ToolsConfig`, a
`shell::zsh_fork_backend` facade, and an opaque unified-exec
spawn-lifecycle hook so unified exec can prepare a wrapped `zsh -c/-lc`
request without storing `EscalationSession` directly in generic
process/runtime code.
- Kept the existing `shell_command` zsh-fork behavior intact on top of
the new bootstrap path. Tool selection is unchanged in this PR: when
`shell_zsh_fork` is enabled, `ShellCommand` still wins over
`exec_command`.
## Verification
- `cargo test -p codex-shell-escalation`
- includes coverage for `start_session_exposes_wrapper_env_overlay`
- includes coverage for `exec_closes_parent_socket_after_shell_spawn`
- includes coverage for
`dropping_session_aborts_intercept_workers_and_kills_spawned_child`
- `cargo test -p codex-core
shell_zsh_fork_prefers_shell_command_over_unified_exec`
- `cargo test -p codex-core --test all
shell_zsh_fork_prompts_for_skill_script_execution`
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/13392).
* #13432
* __->__ #13392
## Summary
Changes the permission profile shape from a bare network boolean to a
nested object.
Before:
```yaml
permissions:
network: true
```
After:
```yaml
permissions:
network:
enabled: true
```
This also updates the shared Rust and app-server protocol types so
`PermissionProfile.network` is no longer `Option<bool>`, but
`Option<NetworkPermissions>` with `enabled: Option<bool>`.
## What Changed
- Updated `PermissionProfile` in `codex-rs/protocol/src/models.rs`:
- `pub network: Option<bool>` -> `pub network:
Option<NetworkPermissions>`
- Added `NetworkPermissions` with:
- `pub enabled: Option<bool>`
- Changed emptiness semantics so `network` is only considered empty when
`enabled` is `None`
- Updated skill metadata parsing to accept `permissions.network.enabled`
- Updated core permission consumers to read
`network.enabled.unwrap_or(false)` where a concrete boolean is needed
- Updated app-server v2 protocol types and regenerated schema/TypeScript
outputs
- Updated docs to mention `additionalPermissions.network.enabled`
Historically, we cloned the Bash repo from
https://github.com/bminor/bash, but for whatever reason, it was removed
at some point.
I had a local clone of it, so I pushed it to
https://github.com/bolinfest/bash so that we could continue running our
CI job. I did this in https://github.com/openai/codex/pull/9563, and as
you can see, I did not tamper with the commit hash we used as the basis
of this build.
Using a personal fork is not great, so this PR changes the CI job to use
what appears to be considered the source of truth for Bash, which is
https://git.savannah.gnu.org/git/bash.git.
Though in testing this out, it appears this Git server does not support
the combination of `git clone --depth 1
https://git.savannah.gnu.org/git/bash` and `git fetch --depth 1 origin
a8a1c2fac029404d3f42cd39f5a20f24b6e4fe4b`, as it fails with the
following error:
```
error: Server does not allow request for unadvertised object a8a1c2fac029404d3f42cd39f5a20f24b6e4fe4b
```
so unfortunately this means that we have to do a full clone instead of a
shallow clone in our CI jobs, which will be a bit slower.
Also updated `codex-rs/shell-escalation/README.md` to reflect this
change.
## Why
Before this change, an escalation approval could say that a command
should be rerun, but it could not carry the sandbox configuration that
should still apply when the escalated command is actually spawned.
That left an unsafe gap in the `zsh-fork` skill path: skill scripts
under `scripts/` that did not declare permissions could be escalated
without a sandbox, and scripts that did declare permissions could lose
their bounded sandbox on rerun or cached session approval.
This PR extends the escalation protocol so approvals can optionally
carry sandbox configuration all the way through execution. That lets the
shell runtime preserve the intended sandbox instead of silently widening
access.
We likely want a single permissions type for this codepath eventually,
probably centered on `Permissions`. For now, the protocol needs to
represent both the existing `PermissionProfile` form and the fuller
`Permissions` form, so this introduces a temporary disjoint union,
`EscalationPermissions`, to carry either one.
Further, this means that today, a skill either:
- does not declare any permissions, in which case it is run using the
default sandbox for the turn
- specifies permissions, in which case the skill is run using that exact
sandbox, which might be more restrictive than the default sandbox for
the turn
We will likely change the skill's permissions to be additive to the
existing permissions for the turn.
## What Changed
- Added `EscalationPermissions` to `codex-protocol` so escalation
requests can carry either a `PermissionProfile` or a full `Permissions`
payload.
- Added an explicit `EscalationExecution` mode to the shell escalation
protocol so reruns distinguish between `Unsandboxed`, `TurnDefault`, and
`Permissions(...)` instead of overloading `None`.
- Updated `zsh-fork` shell reruns to resolve `TurnDefault` at execution
time, which keeps ordinary `UseDefault` commands on the turn sandbox and
preserves turn-level macOS seatbelt profile extensions.
- Updated the `zsh-fork` skill path so a skill with no declared
permissions inherits the conversation's effective sandbox instead of
escalating unsandboxed.
- Updated the `zsh-fork` skill path so a skill with declared permissions
reruns with exactly those permissions, including when a cached session
approval is reused.
## Testing
- Added unit coverage in
`core/src/tools/runtimes/shell/unix_escalation.rs` for the explicit
`UseDefault` / `RequireEscalated` / `WithAdditionalPermissions`
execution mapping.
- Added unit coverage in
`core/src/tools/runtimes/shell/unix_escalation.rs` for macOS seatbelt
extension preservation in both the `TurnDefault` and
explicit-permissions rerun paths.
- Added integration coverage in `core/tests/suite/skill_approval.rs` for
permissionless skills inheriting the turn sandbox and explicit skill
permissions remaining bounded across cached approval reuse.
## Why
In the `shell_zsh_fork` flow, `codex-shell-escalation` receives the
executable path exactly as the shell passed it to `execve()`. That path
is not guaranteed to be absolute.
For commands such as `./scripts/hello-mbolin.sh`, if the shell was
launched with a different `workdir`, resolving the intercepted `file`
against the server process working directory makes policy checks and
skill matching inspect the wrong executable. This change pushes that fix
a step further by keeping the normalized path typed as `AbsolutePathBuf`
throughout the rest of the escalation pipeline.
That makes the absolute-path invariant explicit, so later code cannot
accidentally treat the resolved executable path as an arbitrary
`PathBuf`.
## What Changed
- record the wrapper process working directory as an `AbsolutePathBuf`
- update the escalation protocol so `workdir` is explicitly absolute
while `file` remains the raw intercepted exec path
- resolve a relative intercepted `file` against the request `workdir` as
soon as the server receives the request
- thread `AbsolutePathBuf` through `EscalationPolicy`,
`CoreShellActionProvider`, and command normalization helpers so the
resolved executable path stays type-checked as absolute
- replace the `path-absolutize` dependency in `codex-shell-escalation`
with `codex-utils-absolute-path`
- add a regression test that covers a relative `file` with a distinct
`workdir`
## Verification
- `cargo test -p codex-shell-escalation`
## Why
`codex-shell-escalation` exposed a `codex-core`-specific adapter layer
(`ShellActionProvider`, `ShellPolicyFactory`, and `run_escalate_server`)
that existed only to bridge `codex-core` to `EscalateServer`. That
indirection increased API surface and obscured crate ownership without
adding behavior.
This change moves orchestration into `codex-core` so boundaries are
clearer: `codex-shell-escalation` provides reusable escalation
primitives, and `codex-core` provides shell-tool policy decisions.
Admittedly, @pakrym rightfully requested this sort of cleanup as part of
https://github.com/openai/codex/pull/12649, though this avoids moving
all of `codex-shell-escalation` into `codex-core`.
## What changed
- Made `EscalateServer` public and exported it from `shell-escalation`.
- Removed the adapter layer from `shell-escalation`:
- deleted `shell-escalation/src/unix/core_shell_escalation.rs`
- removed exports for `ShellActionProvider`, `ShellPolicyFactory`,
`EscalationPolicyFactory`, and `run_escalate_server`
- Updated `core/src/tools/runtimes/shell/unix_escalation.rs` to:
- create `Stopwatch`/cancellation in `codex-core`
- instantiate `EscalateServer` directly
- implement `EscalationPolicy` directly on `CoreShellActionProvider`
Net effect: same escalation flow with fewer wrappers and a smaller
public API.
## Verification
- Manually reviewed the old vs. new escalation call flow to confirm
timeout/cancellation behavior and approval policy decisions are
preserved while removing wrapper types.
## Why
This PR switches the `shell_command` zsh-fork path over to
`codex-shell-escalation` so the new shell tool can use the shared
exec-wrapper/escalation protocol instead of the `zsh_exec_bridge`
implementation that was introduced in
https://github.com/openai/codex/pull/12052. `zsh_exec_bridge` relied on
UNIX domain sockets, which is not as tamper-proof as the FD-based
approach in `codex-shell-escalation`.
## What Changed
- Added a Unix zsh-fork runtime adapter in `core`
(`core/src/tools/runtimes/shell/unix_escalation.rs`) that:
- runs zsh-fork commands through
`codex_shell_escalation::run_escalate_server`
- bridges exec-policy / approval decisions into `ShellActionProvider`
- executes escalated commands via a `ShellCommandExecutor` that calls
`process_exec_tool_call`
- Updated `ShellRuntime` / `ShellCommandHandler` / tool spec wiring to
select a `shell_command` backend (`classic` vs `zsh-fork`) while leaving
the generic `shell` tool path unchanged.
- Removed the `zsh_exec_bridge`-based session service and deleted
`core/src/zsh_exec_bridge/mod.rs`.
- Moved exec-wrapper entrypoint dispatch to `arg0` by handling the
`codex-execve-wrapper` arg0 alias there, and removed the old
`codex_core::maybe_run_zsh_exec_wrapper_mode()` hooks from `cli` and
`app-server` mains.
- Added the needed `codex-shell-escalation` dependencies for `core` and
`arg0`.
## Tests
- `cargo test -p codex-core
shell_zsh_fork_prefers_shell_command_over_unified_exec`
- `cargo test -p codex-app-server turn_start_shell_zsh_fork --
--nocapture`
- verifies zsh-fork command execution and approval flows through the new
backend
- includes subcommand approve/decline coverage using the shared zsh
DotSlash fixture in `app-server/tests/suite/zsh`
- To test manually, I added the following to `~/.codex/config.toml`:
```toml
zsh_path = "/Users/mbolin/code/codex3/codex-rs/app-server/tests/suite/zsh"
[features]
shell_zsh_fork = true
```
Then I ran `just c` to run the dev build of Codex with these changes and
sent it the message:
```
run `echo $0`
```
And it replied with:
```
echo $0 printed:
/Users/mbolin/code/codex3/codex-rs/app-server/tests/suite/zsh
In this tool context, $0 reflects the script path used to invoke the shell, not just zsh.
```
so the tool appears to be wired up correctly.
## Notes
- The zsh subcommand-decline integration test now uses `rm` under a
`WorkspaceWrite` sandbox. The previous `/usr/bin/true` scenario is
auto-allowed by the new `shell-escalation` policy path, which no longer
produces subcommand approval prompts.
## Why
After removing `exec-server`, the next step is to wire a new shell tool
to `codex-rs/shell-escalation` directly.
That is blocked while `codex-shell-escalation` depends on `codex-core`,
because the new integration would require `codex-core` to depend on
`codex-shell-escalation` and create a dependency cycle.
This change ports the reusable pieces from the earlier prep work, but
drops the old compatibility shim because `exec-server`/MCP support is
already gone.
## What Changed
### Decouple `shell-escalation` from `codex-core`
- Introduce a crate-local `SandboxState` in `shell-escalation`
- Introduce a `ShellCommandExecutor` trait so callers provide process
execution/sandbox integration
- Update `EscalateServer::exec(...)` and `run_escalate_server(...)` to
use the injected executor
- Remove the direct `codex_core::exec::process_exec_tool_call(...)` call
from `shell-escalation`
- Remove the `codex-core` dependency from `codex-shell-escalation`
### Restore reusable policy adapter exports
- Re-enable `unix::core_shell_escalation`
- Export `ShellActionProvider` and `ShellPolicyFactory` from
`shell-escalation`
- Keep the crate root API simple (no `legacy_api` compatibility layer)
### Port socket fixes from the earlier prep commit
- Use `socket2::Socket::pair_raw(...)` for AF_UNIX socketpairs and
restore `CLOEXEC` explicitly on both endpoints
- Keep `CLOEXEC` cleared only on the single datagram client FD that is
intentionally passed across `exec`
- Clean up `tokio::AsyncFd::try_io(...)` error handling in the socket
helpers
## Verification
- `cargo shear`
- `cargo clippy -p codex-shell-escalation --tests`
- `cargo test -p codex-shell-escalation`
## Why
We already plan to remove the shell-tool MCP path, and doing that
cleanup first makes the follow-on `shell-escalation` work much simpler.
This change removes the last remaining reason to keep
`codex-rs/exec-server` around by moving the `codex-execve-wrapper`
binary and shared shell test fixtures to the crates/tests that now own
that functionality.
## What Changed
### Delete `codex-rs/exec-server`
- Remove the `exec-server` crate, including the MCP server binary,
MCP-specific modules, and its test support/test suite
- Remove `exec-server` from the `codex-rs` workspace and update
`Cargo.lock`
### Move `codex-execve-wrapper` into `codex-rs/shell-escalation`
- Move the wrapper implementation into `shell-escalation`
(`src/unix/execve_wrapper.rs`)
- Add the `codex-execve-wrapper` binary entrypoint under
`shell-escalation/src/bin/`
- Update `shell-escalation` exports/module layout so the wrapper
entrypoint is hosted there
- Move the wrapper README content from `exec-server` to
`shell-escalation/README.md`
### Move shared shell test fixtures to `app-server`
- Move the DotSlash `bash`/`zsh` test fixtures from
`exec-server/tests/suite/` to `app-server/tests/suite/`
- Update `app-server` zsh-fork tests to reference the new fixture paths
### Keep `shell-tool-mcp` as a shell-assets package
- Update `.github/workflows/shell-tool-mcp.yml` packaging so the npm
artifact contains only patched Bash/Zsh payloads (no Rust binaries)
- Update `shell-tool-mcp/package.json`, `shell-tool-mcp/src/index.ts`,
and docs to reflect the shell-assets-only package shape
- `shell-tool-mcp-ci.yml` does not need changes because it is already
JS-only
## Verification
- `cargo shear`
- `cargo clippy -p codex-shell-escalation --tests`
- `just clippy`
## Why
Shell execution refactoring in `exec-server` had become split between
duplicated code paths, which blocked a clean introduction of the new
reusable shell escalation flow. This commit creates a dedicated
foundation crate so later shell tooling changes can share one
implementation.
## What changed
- Added the `codex-shell-escalation` crate and moved the core escalation
pieces (`mcp` protocol/socket/session flow, policy glue) that were
previously in `exec-server` into it.
- Normalized `exec-server` Unix structure under a dedicated `unix`
module layout and kept non-Unix builds narrow.
- Wired crate/build metadata so `shell-escalation` is a first-class
workspace dependency for follow-on integration work.
## Verification
- Built and linted the stack at this commit point with `just clippy`.
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/12556).
* #12584
* #12583
* __->__ #12556
Michael Bolin ·
Andrei Eternal ·
Charley Cunningham ·
Celia Chen ·