Plan Mode: read-only, iterative planning (Shift+Tab) + configurable planner - #4769
Closed
thomast8 wants to merge 1502 commits into
Closed
Plan Mode: read-only, iterative planning (Shift+Tab) + configurable planner#4769thomast8 wants to merge 1502 commits into
thomast8 wants to merge 1502 commits into
Conversation
Unify the animation in a single code and add the CTRL + . in the onboarding
Only emit a task finished when the compaction comes from a `/compact`
<img width="931" height="438" alt="Screenshot 2025-09-16 at 4 25 19 PM" src="https://github.com/user-attachments/assets/ccfb8df1-feaf-45b4-8f7f-56100de916d5" />
Fix some small references issue. No behavioural change. Just making the code cleaner
openai#3850) ### Why Use `tokio::sync::Mutex` `std::sync::Mutex` are not _async-aware_. As a result, they will block the entire thread instead of just yielding the task. Furthermore they can be poisoned which is not the case of `tokio` Mutex. This allows the Tokio runtime to continue running other tasks while waiting for the lock, preventing deadlocks and performance bottlenecks. In general, this is preferred in async environment
Adds `/review` action in TUI <img width="637" height="370" alt="Screenshot 2025-09-17 at 12 41 19 AM" src="https://github.com/user-attachments/assets/b1979a6e-844a-4b97-ab20-107c185aec1d" />
…penai#3874) Previous to this PR, both of these functions take a single `cwd`: https://github.com/openai/codex/blob/71038381aa0f51aa62e1a2bcc7cbf26a05b141f3/codex-rs/core/src/seatbelt.rs#L19-L25 https://github.com/openai/codex/blob/71038381aa0f51aa62e1a2bcc7cbf26a05b141f3/codex-rs/core/src/landlock.rs#L16-L23 whereas `cwd` and `sandbox_cwd` should be set independently (fixed in this PR). Added `sandbox_distinguishes_command_and_policy_cwds()` to `codex-rs/exec/tests/suite/sandbox.rs` to verify this.
…s in order (openai#3881) Incidentally, we had a test for this in `accepts_multiple_commands_with_allowed_operators()`, but it was verifying the bad behavior. Oops!
Currently, we change the tool description according to the sandbox policy and approval policy. This breaks the cache when the user hits `/approvals`. This PR does the following: - Always use the shell with escalation parameter: - removes `create_shell_tool_for_sandbox` and always uses unified tool via `create_shell_tool` - Reject the func call when the model uses escalation parameter when it cannot.
These are generally useful
Move to using test helper method everywhere.
Hopefully fixes incorrectly showing ^J instead of Shift+Enter in the key hints occasionally.
We currently get information about rate limits in the response headers. We want to forward them to the clients to have better transparency. UI/UX plans have been discussed and this information is needed.
### /limits: show rate limits graph <img width="442" height="287" alt="image" src="https://github.com/user-attachments/assets/3e29a241-a4b0-4df8-bf71-43dc4dd805ca" /> ### Warning on close to rate limits: <img width="507" height="96" alt="image" src="https://github.com/user-attachments/assets/732a958b-d240-4a89-8289-caa92de83537" /> Based on openai#3965
Adds the following options: 1. Review current changes 2. Review a specific commit 3. Review against a base branch (PR style) 4. Custom instructions <img width="487" height="330" alt="Screenshot 2025-09-20 at 2 11 36 PM" src="https://github.com/user-attachments/assets/edb0aaa5-5747-47fa-881f-cc4c4f7fe8bc" /> --- \+ Adds the following UI helpers: 1. Makes list selection searchable 2. Adds navigation to the bottom pane, so you could add a stack of popups 3. Basic custom prompt view
Makes tests shorter
Unify cargo versions at root
The only file to watch is the cargo.toml All the others come from just fix + a few manual small fix The set of rules have been taken from the list of clippy rules arbitrarily while trying to optimise the learning and style of the code while limiting the loss of productivity
defaults to 60sec, overridable with MCP_TOOL_TIMEOUT or on a per-server basis in the config.
Add limits to status <img width="579" height="430" alt="image" src="https://github.com/user-attachments/assets/d3794d92-ffca-47be-8011-b4452223cc89" />
no intended functional change, just simplifying the code.
Backtracking multiple times could drop earlier turns. We now derive the active user-turn positions from the transcript on demand (keying off the latest session header) instead of caching state. This keeps the replayed context intact during repeated edits and adds a regression test.
Moved to /status openai#4053
Adds a "View Stack" to the bottom pane to allow for pushing/popping bottom panels. `esc` will go back instead of dismissing. Benefit: We retain the "selection state" of a parent panel (e.g. the review panel).
Add redundant closure clippy rules and let Codex fix it by minimising FQP
…vers (openai#4689) The `experimental_use_rmcp_client` flag is still useful to: 1. Toggle between stdio clients 2. Enable oauth beacuse we want to land modelcontextprotocol/rust-sdk#469, openai#4677, and binary signing before we enable it by default However, for no-auth http servers, there is only one option so we don't need the flag and it seems to be working pretty well.
## Summary - ensure the issue deduplicator workflow ignores the current issue when listing potential duplicates ## Testing - not run (workflow change) ------ https://chatgpt.com/codex/tasks/task_i_68e03244836c8320a4aa22bfb98fd291
Include information about the action and SDK
fixes an issue when terminals change their color scheme, e.g. dark/light mode, the composer wouldn't update its background color.
Co-authored-by: rakesh <rakesh@openai.com>
It's present on the event, add it to the final result as well.
# External (non-OpenAI) Pull Request Requirements Before opening this Pull Request, please read the dedicated "Contributing" markdown file or your PR may be closed: https://github.com/openai/codex/blob/main/docs/contributing.md If your PR conforms to our contribution guidelines, replace this text with a detailed and high quality description of your changes.
We truncate the output of exec commands to not blow the context window. However, some cases we weren't doing that. This caused reports of people with 76% context window left facing `input exceeded context window` which is weird.
In the past, we were treating `input exceeded context window` as a streaming error and retrying on it. Retrying on it has no point because it won't change the behavior. In this PR, we surface the error to the client without retry and also send a token count event to indicate that the context window is full. <img width="650" height="125" alt="image" src="https://github.com/user-attachments/assets/c26b1213-4c27-4bfc-90f4-51a270a3efd5" />
## Summary This PR is an alternative approach to openai#4711, but instead of changing our storage, parses out shell calls in the client and reserializes them on the fly before we send them out as part of the request. What this changes: 1. Adds additional serialization logic when the ApplyPatchToolType::Freeform is in use. 2. Adds a --custom-apply-patch flag to enable this setting on a session-by-session basis. This change is delicate, but is not meant to be permanent. It is meant to be the first step in a migration: 1. (This PR) Add in-flight serialization with config 2. Update model_family default 3. Update serialization logic to store turn outputs in a structured format, with logic to serialize based on model_family setting. 4. Remove this rewrite in-flight logic. ## Test Plan - [x] Additional unit tests added - [x] Integration tests added - [x] Tested locally
Remove codex-level workingDirectory Throw on turn.failed in `run()` Cleanup readme
When truncating output, add a hint of the total number of lines
## Summary - replace manual event polling loops in several core test suites with the shared wait_for_event helpers - keep prior assertions intact by using closure captures for stateful expectations, including plan updates, patch lifecycles, and review flow checks - rely on wait_for_event_with_timeout where longer waits are required, simplifying timeout handling ## Testing - just fmt ------ https://chatgpt.com/codex/tasks/task_i_68e1d58582d483208febadc5f90dd95e
## Summary Updates tool config for gpt-5-codex ## Test Plan - [x] Ran locally - [x] Updated unit tests
- Add PlanRequested flow for Shift+Tab; block during tasks and keep text. - Default planning to gpt-5 with high effort; make plan model/effort configurable. - Add /plan-model selector with two-step model/effort UI and persistence. - Core config: add plan_model + plan_model_reasoning_effort; add persistence helper. - App protocol/profile structs updated to expose planning model settings. - Footer hint shows Shift+Tab; snapshots updated; tests added/updated. - Replace planning prompt with concise Plan-Only Mode instructions. Refs: feature/plan-shortcut
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #4118
What
/plan-model.update_plantool calls as a checklist in the transcript.Why
How (technical summary)
plan_modeand a Shift+Tab (BackTab) toggle. Enter in plan mode calls a planning path instead of sending a normal turn. Composer remains in plan mode across submissions and is blocked while a task is running.request_plan_turntemporarily switches to a read-only sandbox and (optionally) a planner model/effort for the planning turn, then restores the prior context. It prompts the model to record a concise plan viaupdate_plan.PlanUpdateevents are displayed as a checkbox list (pending, in_progress, completed) with an optional explanation./plan-modelto adjust them.How to use (high level)
/plan-modelto pick a model and reasoning effort.Notes on contributing
docs/contributing.mdand kept the change focused with tests and docs updates. This PR links to and closes the tracked issue.Verification steps
cargo test -p codex-tuithencodex.