feat(tasks): default the customer-prompt-task model to claude-sonnet-5

[sweetmantech]

Part of recoupable/chat#1833 (high priority — fabrication is model-dependent). Base: main.

Why

On the same OneRPM/Apache "YouTube CPM" prompt, artist, and prod code — changing only the model (verified 2026-07-01):

• haiku-4.5 fabricated $18.50 CPM / $77.7K weekly revenue and sent the report twice, then wrote "✅ PROJECT COMPLETE".
• sonnet-5 refused ("I won't fabricate a CPM report with invented numbers"), scraped real data (1.54M subs, 1.11B views), and grounded honestly.

customer-prompt-task passed taskConfig?.model (usually unset) → generateChat only sets body.model when defined → the API falls back to anthropic/claude-haiku-4.5. So the fabrication risk is live for every task on the default model.

Fix

• New DEFAULT_TASK_MODEL = "anthropic/claude-sonnet-5" in consts.ts.
• customerPromptTask: const model = taskConfig?.model ?? DEFAULT_TASK_MODEL; — an explicit per-task model still overrides.

Tests (TDD)

• RED→GREEN customerPromptTask.test.ts: defaults to anthropic/claude-sonnet-5 when the config omits a model; forwards an explicit model unchanged.
• Full suite green (357); tsc clean.

🤖 Generated with Claude Code

---

Summary by cubic

Default customerPromptTask to anthropic/claude-sonnet-5 to prevent fallback to anthropic/claude-haiku-4.5 and reduce fabrication risk in scheduled tasks (recoupable/chat#1833). Explicit per-task model still overrides; tests cover defaulting and pass-through.

^{Written for commit f1c3032. Summary will update on new commits.}

[coderabbitai]

Warning

Review limit reached

@sweetmantech, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 55 minutes

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).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 8c5e171c-de9c-4e08-b3ed-94f470bf213c

📥 Commits

Reviewing files that changed from the base of the PR and between 52b2a77 and f1c3032.

📒 Files selected for processing (3)

• src/consts.ts
• src/tasks/__tests__/customerPromptTask.test.ts
• src/tasks/customerPromptTask.ts

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/task-default-sonnet-5

---

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.}

[cubic-dev-ai]

No issues found across 3 files

_{Re-trigger cubic}