feat(web): edit scheduled task effort and permissions

[kalvenschraut]

Problem

The web and desktop scheduled-task editor (Settings → Scheduled tasks) only had a model picker. A task's reasoning effort and permissions (runtime mode) were fixed when it was created, and an agent can't edit a task whose permissions are broader than its own thread's, so there was no way to change them short of recreating the task. Mobile already offered both controls.

Fix

• An effort picker (TraitsPicker) sits next to the model picker, the same pairing the project-defaults settings use. Effort is kept while the model is unchanged and cleared when you pick a different model or environment. Claude's Ultrathink adds a prefix to the task prompt, as it does in the chat composer.
• A Permissions select lists only the runtime modes the task's provider supports and uses the composer's fallback for a stored mode the provider no longer offers. It is disabled for tasks bound to an existing thread, because those runs keep the thread's permissions.
• The draft now stores modelOptions instead of baseModelSelection. draftModelSelection() builds the saved selection, and it has logic tests.

Verification

• scheduledTasksSettings.logic.test.ts (24 tests), tsc --noEmit for apps/web, and lint on the changed files all pass.
• Codex gpt-6-astra (xhigh) reviewed the change. Round 1 found the two issues fixed in the second commit (Ultrathink did nothing; Permissions had no effect on bound tasks). Round 2 approved with no findings.
• Not checked in a browser.

Model and harness: Claude Opus 5.5 in Claude Code via T3 Code, reviewed by Codex gpt-6-astra.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 40 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Path: .coderabbit.config.ts
• Review profile: CHILL
• Plan: Advanced
• Run ID: 32e0ba08-47ed-4792-adb3-3a9a7af4507d

---

📥 Commits

Reviewing files that changed from the base of the PR and between a849bed and f0f0f7f.

---

📒 Files selected for processing (3)

• apps/web/src/components/settings/ScheduledTasksSettings.tsx
• apps/web/src/components/settings/scheduledTasksSettings.logic.test.ts
• apps/web/src/components/settings/scheduledTasksSettings.logic.ts

---

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	4.9 KiB	−23 B (−0.5%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	−23 B (−1.9%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	20.9 KiB	20.8 KiB	−41 B (−0.2%)	29.3 KiB	✅
Codex	Live turn messages	2	1	−1 (−50.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.2 KiB	21.2 KiB	0 B (0.0%)	29.3 KiB	✅
Claude	Live turn messages	2	2	0 (0.0%)	8	✅

Baseline: a849bed · PR result: f0f0f7f · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.