Make Claude Sonnet 5.5 the Pro default and Opus opt-in only by Programcode · Pull Request #164 · Programcode/DebateOwl

Programcode · 2026-10-22T10:59:29Z

Cost fix 1 of 7. Pro accounts now start on Claude Sonnet 5.5 instead of Claude Opus 5.5, and Opus runs only when someone picks it.

What changes

DEFAULT_PRO_MODEL is now anthropic/claude-sonnet-5.5. New Pro accounts, and Pro accounts that never saved a model, get Sonnet for every tool.
Auto Pro no longer has Opus in its pool. It chooses between Sonnet 5.5 (the default) and GPT-6.1 Sol. When the router is unsure, fails, or names a model outside the pool, it falls back to Sonnet, never Opus.
A saved retired Sonnet ID (claude-sonnet-5, claude-sonnet-4-6) moves to Sonnet 5.5. A saved retired Opus ID (claude-opus-5, claude-opus-4-7) still moves to Opus 5.5, because someone picked Opus on purpose.
Opus 5.5 stays in the catalog and in the selector for Pro accounts.

What this does not change

Existing Pro accounts that already saved Opus 5.5 keep it. The app keeps any saved model that is still in the catalog. The saved row doesn't record whether someone picked Opus or just got it as the old default, so the client can't tell the two apart. Moving those accounts to Sonnet would mean a production data change to user_settings.feature_configs, which this PR doesn't do. Until then, the savings apply to new accounts, accounts that never saved a model, and every Auto Pro request.

This PR doesn't touch the research model, token caps, quotas, prompts, cost logging, or JSON output. Other cost PRs cover those.

Verification (local; GitHub Actions is down repo-wide)

npm run test:edge: 357 passed. The new router test proves that Auto Pro rejects an Opus answer from the router and falls back to Sonnet when the router fails. The handler test now expects Auto Pro to call Sonnet.
npm run verify: lint, typecheck, unit, edge, build, service-worker check and 30 desktop Chromium smoke tests all passed.
verify-debatelab fake stack, recipes/model-defaults on desktop-chromium and mobile-chromium: 3 of 3 passed on each. A new Pro account's user_settings hold Sonnet 5.5 for every tool, the selector shows Claude Sonnet 5.5 with Opus 5.5 still enabled, and a Generate run records model: anthropic/claude-sonnet-5.5 in ai_requests. Basic accounts still start on Haiku 5.5.
Not checked: live OpenRouter calls (live verify pins every call to Muse Spark anyway), staging, or production.

I updated the verify skill's recipe and feature notes, docs/architecture.md and supabase/README.md to match.

🤖 Generated with Claude Code

New Pro accounts start on Sonnet 5.5. Auto Pro routes between Sonnet 5.5 and GPT-6.1 Sol and falls back to Sonnet, so it never picks Opus. Opus runs only when someone selects it. Retired Opus IDs still move to Opus 5.5, since those were explicit picks. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

vercel · 2026-10-22T10:59:31Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
debate-navigator	Ready	Preview	Oct 22, 2026 1:47pm UTC
debatelab-staging	Ready	Preview	Oct 22, 2026 1:47pm UTC

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

coderabbitai · 2026-10-22T10:59:39Z

Warning

Review limit reached

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

Next included review available in 39 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

Source mode: Auto
Repository: Programcode/DebateLab
Branch: cost/sonnet-pro-default
File: .coderabbit.yaml
Commit: 0aad1e4
Layers: Repository YAML (base), Organization UI (inherited)
Review profile: CHILL
Plan: Free
Run ID: 14af0073-e806-448e-aa96-729ec0b15ee4

📥 Commits

Reviewing files that changed from the base of the PR and between 1ad9329 and 0aad1e4.

📒 Files selected for processing: 11

.claude/skills/verify-debatelab/features/README.md
.claude/skills/verify-debatelab/features/case-generation.md
.claude/skills/verify-debatelab/features/model-defaults.md
.claude/skills/verify-debatelab/recipes/model-defaults.spec.ts
docs/architecture.md
src/lib/stores/settings-store.ts
supabase/README.md
supabase/functions/_shared/auto-model-router.test.ts
supabase/functions/_shared/auto-model-router.ts
supabase/functions/_shared/model-catalog.ts
supabase/functions/ai/handler-research.test.ts

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

Note

🎁 Summarized by CodeRabbit Free

Your organization is on the Free plan. CodeRabbit will generate a high-level summary and a walkthrough for each pull request. For a comprehensive line-by-line review, please upgrade your subscription to CodeRabbit Essentials by visiting https://app.coderabbit.ai/settings/billing.

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

Programcode · 2026-10-22T11:15:46Z

The GitHub Actions failures on f31ecb6 did not run the code. Every job stopped within 1 to 2 seconds, and the annotation reads: "The job was not started because recent account payments have failed or your spending limit needs to be increased." It is the repo-wide Actions billing outage. Local results at f31ecb6: npm run test:edge 357 passed; npm run verify passed (lint, typecheck, unit, edge, build, 30 desktop Chromium smoke tests); verify-debatelab recipes/model-defaults passed 3 of 3 on desktop-chromium and on mobile-chromium. That is a local pass, not a CI pass.

Programcode · 2026-10-22T12:14:06Z

Follow-up data migration for accounts that still have Opus 5.5 saved from the old default: #180. It is stacked on this PR and must be applied after this one is live in production.

vercel · 2026-10-22T13:47:05Z

Deployment failed for project debatelab-staging with the following error:

Resource is limited - try again in 24 hours (more than 100, code: "api-deployments-free-per-day").

Learn More: https://vercel.com/adam-gaborjanis-projects?upgradeToPro=build-rate-limit

… (#180) * Make Claude Sonnet 5.5 the Pro default and Opus opt-in only New Pro accounts start on Sonnet 5.5. Auto Pro routes between Sonnet 5.5 and GPT-6.1 Sol and falls back to Sonnet, so it never picks Opus. Opus runs only when someone selects it. Retired Opus IDs still move to Opus 5.5, since those were explicit picks. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Move saved Opus 5.5 settings to Sonnet 5.5 Opus 5.5 was the Pro default, and a saved setting can't tell the default from a pick. This data migration moves every saved Opus 5.5 tool setting to Sonnet 5.5, backs the original rows up in private.opus_default_backup, and skips users it has already migrated. A rollback script restores Opus wherever the migrated Sonnet is still saved. A database test runs both twice against seeded rows. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Cover more rollback and stale-client cases in the Opus migration test The test now seeds a tool already on Sonnet before the migration, which the rollback must keep, and an old updated_at, so it fails if the migration stops bumping it. Clients only save their settings cleanup while updated_at is unchanged, so the bump keeps an older client from writing Opus back. agent-stack.sh test now runs the script too, like npm run test:db. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

Copilot AI balanced review requested due to automatic review settings October 22, 2026 10:59

Copilot AI reviewed Oct 22, 2026

vercel Bot deployed to Preview – debatelab-staging October 22, 2026 11:01

View deployment

vercel Bot deployed to Preview – debate-navigator October 22, 2026 11:02

View deployment

Programcode mentioned this pull request Oct 22, 2026

Move saved Opus 5.5 settings to Sonnet 5.5 (data migration, after #164) #180

Merged

Programcode mentioned this pull request Oct 22, 2026

Fix the 11 bugs from today's QA pass (H1, M1–M5, L1–L5) #184

Merged

Merge remote-tracking branch 'origin/main' into s164

0aad1e4

Programcode merged commit 988b870 into main Oct 22, 2026
4 of 17 checks passed

vercel Bot deployed to Preview – debate-navigator October 22, 2026 13:47

View deployment

Programcode mentioned this pull request Oct 22, 2026

Run Claude only on the Anthropic BYOK key, with no fallback provider #185

Merged

Navigation Menu

Repository navigation

Make Claude Sonnet 5.5 the Pro default and Opus opt-in only - #164

Make Claude Sonnet 5.5 the Pro default and Opus opt-in only#164
Programcode merged 2 commits into
mainfrom
cost/sonnet-pro-default

Programcode commented Oct 22, 2026

Uh oh!

vercel Bot commented Oct 22, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

coderabbitai Bot commented Oct 22, 2026 •

edited

Loading

Review limit reached

Uh oh!

Programcode commented Oct 22, 2026

Uh oh!

Programcode commented Oct 22, 2026

Uh oh!

vercel Bot commented Oct 22, 2026

Uh oh!

Uh oh!

This branch was successfully deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Programcode commented Oct 22, 2026

What changes

What this does not change

Verification (local; GitHub Actions is down repo-wide)

Uh oh!

vercel Bot commented Oct 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot commented Oct 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review limit reached

Uh oh!

Programcode commented Oct 22, 2026

Uh oh!

Programcode commented Oct 22, 2026

Uh oh!

vercel Bot commented Oct 22, 2026

Uh oh!

Uh oh!

This branch was successfully deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

vercel Bot commented Oct 22, 2026 •

edited

Loading

coderabbitai Bot commented Oct 22, 2026 •

edited

Loading