[waleedlatif1]

Summary

• Free-plan spend past the one-time allowance is never billed, but runs could overshoot it: the verdict returned after each Mothership cost callback came from a cached admission (free payers are always "near limit", so up to a 10 s window), and an unreadable or slow standing read answered not-exceeded
• After a cost callback, a free payer's standing is now read fresh from the ledger (checkAttributedUsageLimits, the function the cache wraps; the direct-v1 path skips its verdict cache and shared in-flight read), so the callback that crosses the limit pauses the run
• Free payers fail closed when standing can't be read — update-cost (unknown or the 1 s budget), the provider mid-run check, the copilot continuation re-check, and Mothership continuation admission all stop/refuse a free run with the usage card. Paid plans and orgs keep today's fail-open behaviour so a database blip never ends a paying run
• isFreeRunPayer decides free from what the run was admitted with (attributed payerSubscription; a direct-v1 decision without payerSubscriptionId), so no new query; errors land on the paid (status-quo) side
• Default free-plan concurrency 10 → 2 (docs table updated); admission caching for workflow executions is unchanged

Type of Change

• Bug fix

Testing

• app/api/billing/update-cost/route.integration.ts (real Postgres): free direct-v1 and free attributed payers are paused when a second callback crosses the limit inside the cache window; a pro payer in the same scenario keeps the cached answer. Red with the fresh-read bypass disabled, green restored
• Unit: update-cost/route.test.ts (fresh read, unreadable ledger, failed gate read, timeout → exceeded for free; blocked → not exceeded; paid unchanged), providers/run-usage-limit.test.ts, copilot/api-keys/validate/route.test.ts, mothership/request/lifecycle/admission.test.ts — each new free-payer guard shown red with the guard reverted
• lib/billing + touched route/provider/lifecycle suites green; biome, check:test-patterns, check:comment-hygiene, check:unused-exports, check:file-names, check:boundaries
• Full type-check / audits / suites left to CI

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing (new tests pass the test-audit authoring gate)
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

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

1 Skipped Deployment

Project	Deployment	Actions	Updated
docs	Ignored	Preview	Oct 30, 2026 1:34am UTC

[greptile-apps]

Confidence Score: 5/5

[Medium impact] The PR appears safe to merge; no new actionable issues were found.

Summary

Free runs now read their latest spend after cost callbacks and stop when their allowance is spent or cannot be checked. Paid runs keep their existing behavior.

• Free runs pause when the allowance is reached or cannot be checked.
• Free accounts now have a default cap of two concurrent workflow executions.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart TD
  A[Cost recorded] --> B{Free payer?}
  B -->|Yes| C[Read latest spend]
  B -->|No| D[Use existing cached check]
  C --> E{Standing}
  E -->|Within allowance| F[Continue]
  E -->|Exceeded or unreadable| G[Pause with usage card]
  D --> H{Standing}
  H -->|Exceeded| G
  H -->|Within or unreadable| F

Loading

_{Reviews (3) · Last reviewed commit: "fix(billing): bound free payers' fresh u..." · Reviewed by Greptile}

[cubic-dev-ai]

All reported issues were addressed across 12 files

<sub>This PR changes payment, billing, or quota logic. Ultrareviews find 2.4x more serious bugs than standard reviews. Comment @cubic-dev-ai ultrareview to run one.

Turn on auto-fix | Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic-dev-ai review this PR

[cubic-dev-ai]

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

All reported issues were addressed across 13 files

<sub>Reply with feedback, questions, or to request a fix.

Turn on auto-fix | Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic-dev-ai review this PR

[cubic-dev-ai]

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 14 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

_{Turn on auto-fix | Re-trigger cubic}