feat(autonomy): planner gate for non-Claude providers at level off (P3.T2)#392
Conversation
|
Bugbot is not enabled for your account, so this pull request was not reviewed. Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs. |
|
Warning Review limit reached
Next review available in: 48 minutes Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. How can I continue?After more reviews become available, a review can be triggered using the 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 configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Run ID: 📒 Files selected for processing (3)
✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
…3.T2) AUTO_CODE_AUTONOMY=off now blocks non-Claude planner sessions with a clear ProviderError before the session exists. The check runs in create_planner_session on the effective provider (after runner routing), so a route to a different engine is enforced too. Claude sessions keep working at every level; claude/safe/bold keep non-Claude planners available (default level is claude, so unset env stays permissive). Tests: 10 new (helper matrix incl. default level + provider naming in the error; end-to-end through create_planner_session with a provider stub proving create_session is never reached when gated). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Signed-off-by: Oleg Miagkov <mrobenner@gmail.com>
OBenner
force-pushed
the
claude/p3-planner-autonomy-gate
branch
from
1d932eb to
c1c0fbe
Compare
|
1 participant
What
Roadmap P3·T2:
AUTO_CODE_AUTONOMY=offmust block the non-Claude planner. The coder path already respects the autonomy level (P3·T1); the planner had no gate at all — atoff, a direct-API planner would still spin up and writeimplementation_plan.json.agents/planner.py— new_assert_planner_autonomy_allows(provider_name)called fromcreate_planner_session()after runner routing, on the effective provider (so a route to a different engine is enforced too). At leveloff+ non-Claude provider it raisesProviderErrorwith an actionable message before the session exists.claude/safe/boldkeep non-Claude planners available. Unset env resolves to the default level (claude) and stays permissive.Acceptance (roadmap P3·T2)
Verified end-to-end via a provider stub: at
offthe gate raises andcreate_sessionis never reached; atsafethe non-Claude session is created; atoffthe Claude session is created.Tests
tests/test_planner_autonomy_gate.py(helper matrix incl. default-level behavior and provider naming in the error; 3 end-to-end черезcreate_planner_session).test_planner.py,test_autonomy_level.py— 60 passed total.🤖 Generated with Claude Code