fix(knowledge): increase hosted rerank capacity

[TheodoreSpeaks]

Summary

• Increase Sim-hosted Cohere rerank admission from 60 to 600 requests/minute and burst 2 to 16, keeping the 30-second timeout and shared per-key cooldowns.
• Preserve BYOK, self-hosted, OCR, and embedding defaults. Keep the shared RPM override and add a hosted-specific override with precedence; invalid hosted budgets fail before admission.

Type of Change

• Bug fix

Testing

• 52 focused tests passed, covering credential routing, overrides, shared cooldowns, deadlines, and sustained throughput through the PostgreSQL adapter with mocked persistence.
• App type-check, repository lint, all 46 audits, API validation, block registry checks, and docs manifest validation passed.

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• 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	Skipped		Sep 17, 2026 2:24am UTC

[greptile-apps]

Confidence Score: 5/5

The PR appears safe to merge, with no concrete correctness, security, or repository-rule violations identified.

Summary

This PR increases admission capacity for Sim-hosted Cohere reranking while retaining existing limits for BYOK, self-hosted, OCR, embedding, and other rerank credentials.

• Adds a hosted-specific rerank RPM override with precedence over the shared override.
• Classifies resolved Cohere credentials before selecting the admission budget.
• Raises hosted capacity to a 16-request burst and a sustained 600 requests per minute by default.
• Adds coverage for credential routing, configuration validation, cooldown sharing, deadlines, and token-bucket throughput.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart LR
  Request[Rerank request] --> Resolve[Resolve Cohere credential]
  Resolve --> BYOK{BYOK credential?}
  BYOK -->|Yes| Standard[Standard rerank budget]
  BYOK -->|No| Hosted{Hosted Sim runtime and platform key?}
  Hosted -->|No| Standard
  Hosted -->|Yes| Expanded[Hosted rerank budget]
  Standard --> Admission[Per-key admission bucket and cooldown]
  Expanded --> Admission
  Admission --> Cohere[Cohere rerank API]

Loading

_{Reviews (1) · Last reviewed commit: "fix(knowledge): increase hosted rerank c..."}

[cubic-dev-ai]

No issues found across 5 files

Confidence score: 5/5

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

_{Re-trigger cubic}