Commit Graph
24 Commits
Author SHA1 Message Date
kevin zhao bf0ba87200 . 2025-12-03 20:06:46 -08:00
kevin zhao be1a90bb6e add back line 2025-12-03 20:04:56 -08:00
kevin zhao 4563dd0a17 Add approval allow-prefix flow in core and tui
Add explicit prefix-approval decision and wire it through execpolicy/UI snapshots

update doc

mutating in memory policy instead of reloading

using RW locks

clippy

refactor: adding allow_prefix into ApprovedAllowPrefix

fmt

do not send allow_prefix if execpolicy is disabled

moving args around

cleanup exec_policy getters

undo diff

fixing rw lock bug causing tui to hang

updating phrasing

integration test

.

fix compile

fix flaky test

fix compile error

running test with single thread

fixup allow_prefix_if_applicable

fix formatting

fix approvals test

only cloning when needed

docs

add docstring

fix rebase bug

fixing rebase issues

Revert "fixing rebase issues"

This reverts commit 79ce7e1f2fc0378c2c0b362408e2e544566540fd.

fix rebase errors
2025-12-03 20:04:56 -08:00
Michael Bolin ec93b6daf3 chore: make create_approval_requirement_for_command an async fn (#7501)
I think this might help with https://github.com/openai/codex/pull/7033
because `create_approval_requirement_for_command()` will soon need
access to `Session.state`, which is a `tokio::sync::Mutex` that needs to
be accessed via `async`.
2025-12-02 15:01:15 -08:00
jif-oai 72b95db12f feat: intercept apply_patch for unified_exec (#7446) 2025-12-02 17:54:02 +00:00
Michael Bolin f56d1dc8fc feat: update process_exec_tool_call() to take a cancellation token (#6972)
This updates `ExecParams` so that instead of taking `timeout_ms:
Option<u64>`, it now takes a more general cancellation mechanism,
`ExecExpiration`, which is an enum that includes a
`Cancellation(tokio_util::sync::CancellationToken)` variant.

If the cancellation token is fired, then `process_exec_tool_call()`
returns in the same way as if a timeout was exceeded.

This is necessary so that in #6973, we can manage the timeout logic
external to the `process_exec_tool_call()` because we want to "suspend"
the timeout when an elicitation from a human user is pending.








---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with [ReviewStack](https://reviewstack.dev/openai/codex/pull/6972).
* #7005
* #6973
* __->__ #6972
2025-11-20 16:29:57 -08:00
pakrym-oai 30ca89424c Always fallback to real shell (#6953)
Either cmd.exe or `/bin/sh`.
2025-11-20 10:58:46 -08:00
zhao-oaiandMichael Bolin 65c13f1ae7 execpolicy2 core integration (#6641)
This PR threads execpolicy2 into codex-core.

activated via feature flag: exec_policy (on by default)

reads and parses all .codexpolicy files in `codex_home/codex`

refactored tool runtime API to integrate execpolicy logic

---------

Co-authored-by: Michael Bolin <mbolin@openai.com>
2025-11-19 16:50:43 -08:00
pakrym-oai ee0484a98c shell_command returns freeform output (#6860)
Instead of returning structured out and then re-formatting it into
freeform, return the freeform output from shell_command tool.

Keep `shell` as the default tool for GPT-5.
2025-11-18 23:38:43 -08:00
zhao-oai e9e644a119 fixing localshell tool calls (#6823)
- Local-shell tool responses were always tagged as
`ExecCommandSource::UserShell` because handler would call
`run_exec_like` with `is_user_shell_cmd` set to true.
- Treat `ToolPayload::LocalShell` the same as other model generated
shell tool calls by deleting `is_user_shell_cmd` from `run_exec_like`
(since actual user shell commands follow a separate code path)
2025-11-18 17:28:26 +00:00
jif-oai 63c8c01f40 feat: better UI for unified_exec (#6515)
<img width="376" height="132" alt="Screenshot 2025-11-12 at 17 36 22"
src="https://github.com/user-attachments/assets/ce693f0d-5ca0-462e-b170-c20811dcc8d5"
/>
2025-11-14 16:31:12 +01:00
pakrym-oai 7b027e7536 Revert "Revert "Overhaul shell detection and centralize command generation for unified exec"" (#6607)
Reverts openai/codex#6606
2025-11-13 16:45:17 -08:00
pakrym-oai e6995174c1 Revert "Overhaul shell detection and centralize command generation for unified exec" (#6606)
Reverts openai/codex#6577
2025-11-13 08:43:00 -08:00
pakrym-oaiandPavel Krymets d28e912214 Overhaul shell detection and centralize command generation for unified exec (#6577)
This fixes command display for unified exec. All `cd`s and `ls`es are
now parsed.

<img width="452" height="237" alt="image"
src="https://github.com/user-attachments/assets/ce92d81f-f74c-485a-9b34-1eaa29290ec6"
/>

Deletes a ton of tests that were doing nothing from shell.rs.

---------

Co-authored-by: Pavel Krymets <pavel@krymets.com>
2025-11-13 08:28:09 -08:00
jif-oai e00eb50db3 feat: only wait for mutating tools for ghost commit (#6534) 2025-11-12 18:16:32 +00:00
Michael Bolin 29364f3a9b feat: shell_command tool (#6510)
This adds support for a new variant of the shell tool behind a flag. To
test, run `codex` with `--enable shell_command_tool`, which will
register the tool with Codex under the name `shell_command` that accepts
the following shape:

```python
{
  command: str
  workdir: str | None,
  timeout_ms: int | None,
  with_escalated_permissions: bool | None,
  justification: str | None,
}
```

This is comparable to the existing tool registered under
`shell`/`container.exec`. The primary difference is that it accepts
`command` as a `str` instead of a `str[]`. The `shell_command` tool
executes by running `execvp(["bash", "-lc", command])`, though the exact
arguments to `execvp(3)` depend on the user's default shell.

The hypothesis is that this will simplify things for the model. For
example, on Windows, instead of generating:

```json
{"command": ["pwsh.exe", "-NoLogo", "-Command", "ls -Name"]}
```

The model could simply generate:

```json
{"command": "ls -Name"}
```

As part of this change, I extracted some logic out of `user_shell.rs` as
`Shell::derive_exec_args()` so that it can be reused in
`codex-rs/core/src/tools/handlers/shell.rs`. Note the original code
generated exec arg lists like:

```javascript
["bash", "-lc", command]
["zsh", "-lc", command]
["pwsh.exe", "-NoProfile", "-Command", command]
```

Using `-l` for Bash and Zsh, but then specifying `-NoProfile` for
PowerShell seemed inconsistent to me, so I changed this in the new
implementation while also adding a `use_login_shell: bool` option to
make this explicit. If we decide to add a `login: bool` to
`ShellCommandToolCallParams` like we have for unified exec:


https://github.com/openai/codex/blob/807e2c27f0a9f2e85c50e7e6df5533f0d9b853c7/codex-rs/core/src/tools/handlers/unified_exec.rs#L33-L34

Then this should make it straightforward to support.
2025-11-12 08:18:57 -08:00
Abhishek Bhardwaj 89591e4246 feature: Add "!cmd" user shell execution (#2471)
feature: Add "!cmd" user shell execution

This change lets users run local shell commands directly from the TUI by
prefixing their input with ! (e.g. !ls). Output is truncated to keep the
exec cell usable, and Ctrl-C cleanly
  interrupts long-running commands (e.g. !sleep 10000).

**Summary of changes**

- Route Op::RunUserShellCommand through a dedicated UserShellCommandTask
(core/src/tasks/user_shell.rs), keeping the task logic out of codex.rs.
- Reuse the existing tool router: the task constructs a ToolCall for the
local_shell tool and relies on ShellHandler, so no manual MCP tool
lookup is required.
- Emit exec lifecycle events (ExecCommandBegin/ExecCommandEnd) so the
TUI can show command metadata, live output, and exit status.

**End-to-end flow**

  **TUI handling**

1. ChatWidget::submit_user_message (TUI) intercepts messages starting
with !.
2. Non-empty commands dispatch Op::RunUserShellCommand { command };
empty commands surface a help hint.
3. No UserInput items are created, so nothing is enqueued for the model.

  **Core submission loop**
4. The submission loop routes the op to handlers::run_user_shell_command
(core/src/codex.rs).
5. A fresh TurnContext is created and Session::spawn_user_shell_command
enqueues UserShellCommandTask.

  **Task execution**
6. UserShellCommandTask::run emits TaskStartedEvent, formats the
command, and prepares a ToolCall targeting local_shell.
  7. ToolCallRuntime::handle_tool_call dispatches to ShellHandler.

  **Shell tool runtime**
8. ShellHandler::run_exec_like launches the process via the unified exec
runtime, honoring sandbox and shell policies, and emits
ExecCommandBegin/End.
9. Stdout/stderr are captured for the UI, but the task does not turn the
resulting ToolOutput into a model response.

  **Completion**
10. After ExecCommandEnd, the task finishes without an assistant
message; the session marks it complete and the exec cell displays the
final output.

  **Conversation context**

- The command and its output never enter the conversation history or the
model prompt; the flow is local-only.
  - Only exec/task events are emitted for UI rendering.

**Demo video**


https://github.com/user-attachments/assets/fcd114b0-4304-4448-a367-a04c43e0b996
2025-10-29 00:31:20 -07:00
Gabriel Peal b0bdc04c30 [MCP] Render MCP tool call result images to the model (#5600)
It's pretty amazing we have gotten here without the ability for the
model to see image content from MCP tool calls.

This PR builds off of 4391 and fixes #4819. I would like @KKcorps to get
adequete credit here but I also want to get this fix in ASAP so I gave
him a week to update it and haven't gotten a response so I'm going to
take it across the finish line.


This test highlights how absured the current situation is. I asked the
model to read this image using the Chrome MCP
<img width="2378" height="674" alt="image"
src="https://github.com/user-attachments/assets/9ef52608-72a2-4423-9f5e-7ae36b2b56e0"
/>

After this change, it correctly outputs:
> Captured the page: image dhows a dark terminal-style UI labeled
`OpenAI Codex (v0.0.0)` with prompt `model: gpt-5-codex medium` and
working directory `/codex/codex-rs`
(and more)  

Before this change, it said:
> Took the full-page screenshot you asked for. It shows a long,
horizontally repeating pattern of stylized people in orange, light-blue,
and mustard clothing, holding hands in alternating poses against a white
background. No text or other graphics-just rows of flat illustration
stretching off to the right.

Without this change, the Figma, Playwright, Chrome, and other visual MCP
servers are pretty much entirely useless.

I tested this change with the openai respones api as well as a third
party completions api
2025-10-27 17:55:57 -04:00
jif-oai 6745b12427 chore: testing on apply_path (#5557) 2025-10-23 17:00:48 +01:00
jif-oai 8e291a1706 chore: clean handle_container_exec_with_params (#5516)
Drop `handle_container_exec_with_params` to have simpler and more
straight forward execution path
2025-10-23 09:24:01 +01:00
pakrym-oai 789e65b9d2 Pass TurnContext around instead of sub_id (#5421)
Today `sub_id` is an ID of a single incoming Codex Op submition. We then
associate all events triggered by this operation using the same
`sub_id`.

At the same time we are also creating a TurnContext per submission and
we'd like to start associating some events (item added/item completed)
with an entire turn instead of just the operation that started it.

Using turn context when sending events give us flexibility to change
notification scheme.
2025-10-21 08:04:16 -07:00
jif-oai 5e4f3bbb0b chore: rework tools execution workflow (#5278)
Re-work the tool execution flow. Read `orchestrator.rs` to understand
the structure
2025-10-20 20:57:37 +01:00
jif-oai dc3c6bf62a feat: parallel tool calls (#4663)
Add parallel tool calls. This is configurable at model level and tool
level
2025-10-05 16:10:49 +00:00
jif-oai 33d3ecbccc chore: refactor tool handling (#4510)
# Tool System Refactor

- Centralizes tool definitions and execution in `core/src/tools/*`:
specs (`spec.rs`), handlers (`handlers/*`), router (`router.rs`),
registry/dispatch (`registry.rs`), and shared context (`context.rs`).
One registry now builds the model-visible tool list and binds handlers.
- Router converts model responses to tool calls; Registry dispatches
with consistent telemetry via `codex-rs/otel` and unified error
handling. Function, Local Shell, MCP, and experimental `unified_exec`
all flow through this path; legacy shell aliases still work.
- Rationale: reduce per‑tool boilerplate, keep spec/handler in sync, and
make adding tools predictable and testable.

Example: `read_file`
- Spec: `core/src/tools/spec.rs` (see `create_read_file_tool`,
registered by `build_specs`).
- Handler: `core/src/tools/handlers/read_file.rs` (absolute `file_path`,
1‑indexed `offset`, `limit`, `L#: ` prefixes, safe truncation).
- E2E test: `core/tests/suite/read_file.rs` validates the tool returns
the requested lines.

## Next steps:
- Decompose `handle_container_exec_with_params` 
- Add parallel tool calls
2025-10-03 13:21:06 +01:00