Repository navigation
🤖 refactor: resolve the workspace reasoning mode from metadata and an unsent pick - #5904
ibetitsmike wants to merge 6 commits into
Conversation
…nsent pick The composer, sends, palette and plan handoff read the reasoning mode through the resolver, and a reasoning pick stays in memory until the next send saves it. Agent switches, the plan handoff and workspace creation no longer write a reasoning mode, and the unused metadata and composer-model helpers go away. Seeding of the old key stays until a later commit deletes it. Stack-Layer: A2e
Stack-Layer: A2e
ibetitsmike
changed the title
Stack-Layer: A2e
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a7f6bc0281
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Stack-Layer: A2e
Stack-Layer: A2e
Stack-Layer: A2e
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 69adb4311d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
1 participant
Summary
A workspace's reasoning mode (standard or Pro) now comes from the resolver, the same way as the model and thinking level: the unsent pick, then the agent's saved settings, then the agent's configured default. A saved per-agent entry owns the choice, so an entry without a mode means standard.
ThinkingProvider, the command palette,getSendOptionsFromStorageand the plan handoff read it there instead of thereasoningMode:{workspace}key.A reasoning pick in a workspace composer or the palette is recorded as an unsent pick and no longer writes
reasoningMode:{workspace}; it still also goes into the per-agent settings cache until #5906.WorkspaceModeAISyncand the plan handoff stop resolving and writing a reasoning mode on agent switches. Workspace creation stops copying the project's reasoning mode into the new workspace's key and the per-agent cache; the creation request still sends it to the backend. With this,ThinkingProviderno longer readsWorkspaceContext, and the unused helpersgetWorkspaceAiSettingsFromMetadataandresolveEffectiveComposerModelare deleted. Metadata seeding still writes the old key until #5907.Stack
Milestone M4 of the server-side settings stack, on top of M3 (#5869 to #5879): per-workspace AI picks (model, thinking level, reasoning mode, agent and Auto routing) are stored in workspace metadata in
config.jsoninstead of per-browser localStorage keys, so every browser shows the same picks. An unsent picker change stays in memory in its tab until a send saves it, and a reload before sending drops it, as the plan accepts. Layer 11 of 20.Testing
ThinkingContext.test.tsx): a corrupt mode in the saved settings shows standard and a saved Pro shows Pro. It used to seed the localStorage key.CoderGatewaystory inApp.astraPro.stories.tsxdrops its check of the removed key and keeps the checks on the Pro button.make static-checkand the touched test suites pass at the top of the stack; CI checks each layer on its own.Review record
efe7e29096, because a fork without a message dropped the source's unsent picks (fixed by 🤖 feat: fork a workspace with the source composer's unsent AI picks #5897, 🤖 refactor: resolve the workspace agent from metadata and an unsent pick #5908 and 🤖 refactor: resolve auto routing from the unsent pick, saved flags and agent defaults #5909); everything else passed. Round 2: PASS on80f2b5379e: forks from all five entry points keep the unsent picks, startup removes the retired keys, and the round 1 checks still pass; the plan handoff was not exercised because the mock model cannot propose a plan. The Codex round 1 fixes made since then have no remote UAT yet; that is left to the next UAT round on 🤖 refactor: keep preferences and workspace AI picks on the server #5916.a7f6bc0281: 3 findings: 2 fixed by 3e0b705 (🤖 refactor: resolve the workspace model from metadata and an unsent pick #5901), 1 declined with the reason in its thread. Mike asked to land the stack as one PR, so further review runs on 🤖 refactor: keep preferences and workspace AI picks on the server #5916.efe7e29096, probe edits in production code).