v0.8.56: opus 5.5, jev, gpt 6 sol, ui consolidation, usage entry

[waleedlatif1]

• feat(providers): add Claude Opus 5.5 (feat(providers): add Claude Opus 5.5 #8141)
• fix(mcp): list servers without Credential Group access (fix(mcp): list servers without Credential Group access #8143)
• refactor(ui): make measurement styles explicit and reuse preview white (refactor(ui): make measurement styles explicit and reuse preview white #8142)
• fix(organizations): respect inherited access when inviting members (fix(organizations): respect inherited access when inviting members #8131)
• improvement(ui): use chips for table job actions (improvement(ui): use chips for table job actions #8082)
• improvement(ui): share public chat composer actions (improvement(ui): share public chat composer actions #8075)
• improvement(ui): share expandable home sections (improvement(ui): share expandable home sections #8067)
• fix(access): align invitations and request navigation (fix(access): align invitations and request navigation #8145)
• refactor(ui): narrow local style forwarding and preview geometry (refactor(ui): narrow local style forwarding and preview geometry #8144)
• fix(billing): let a slow ledger read finish instead of blocking execution at the singleflight default (fix(billing): let a slow ledger read finish instead of blocking execution at the singleflight default #8146)
• feat(models): add GPT-6 Sol and Luna (feat(models): add GPT-6 Sol and Luna #8147)
• fix(billing): bound every ledger aggregate at the database and size the gate deadline for two (fix(billing): bound every ledger aggregate at the database and size the gate deadline for two #8148)
• feat(jev): add TypeSafe evaluation models to Agent (feat(jev): add TypeSafe evaluation models to Agent #8103)
• fix(oauth): rotate a refresh chain only from the token it started from and size the refresh lease for every provider (fix(oauth): rotate a refresh chain only from the token it started from and size the refresh lease for every provider #8152)
• improvement(settings): consolidate access requests in settings (improvement(settings): consolidate access requests in settings #8154)
• fix(billing): keep archived selected workspaces in the invitation applied-state check (fix(billing): keep archived selected workspaces in the invitation applied-state check #8159)
• fix(knowledge): unschedule a connector whose credential the source rejected and prompt to reconnect (fix(knowledge): unschedule a connector whose credential the source rejected and prompt to reconnect #8158)

[vercel]

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

1 Skipped Deployment

Project	Deployment	Actions	Updated
docs	Skipped		Sep 22, 2026 9:03pm UTC

[greptile-apps]

Confidence Score: 5/5

The PR appears safe to merge; no concrete behavioral, security, or repository-rule failures remain.

Summary

This PR combines provider and model catalog expansion with a native TypeSafe Jev evaluation path, billing-ledger timeout hardening, invitation and access-request corrections, MCP catalog authorization changes, and several shared UI refactors.

• Adds Claude Opus 5.5, GPT-6 Sol/Luna, and TypeSafe Jev models with hosted/BYOK credential handling.
• Adds typed Jev state/questions, structured answers, transport validation, usage accounting, and Agent-block integration.
• Bounds billing-ledger aggregates at the database and aligns usage-gate deadlines with those bounds.
• Reconciles inherited organization access during invitations and introduces organization-scoped access-request navigation.
• Lists only workspace-granted managed MCP connectors while retaining execution-time credential checks.
• Consolidates repeated chat, home-section, table-action, and preview styling components.

Diagram

sequenceDiagram
  participant Workflow as Agent workflow
  participant Executor as AgentBlockHandler
  participant Provider as Provider boundary
  participant TypeSafe as TypeSafe Jev API
  participant Billing as Usage accounting

  Workflow->>Executor: model + evaluationState + evaluationQuestions
  Executor->>Executor: Resolve references and select evaluation modality
  Executor->>Provider: Jev evaluation request
  Provider->>Provider: Resolve workspace/org BYOK or hosted key
  Provider->>TypeSafe: POST state and typed questions
  TypeSafe-->>Provider: answers, resolved model, token usage
  Provider-->>Executor: content + structured answers + usage/cost
  Executor-->>Workflow: Agent outputs
  Executor->>Billing: Record billable hosted usage

Loading

_{Reviews (1) · Last reviewed commit: "feat(jev): add TypeSafe evaluation model..."}

[cubic-dev-ai]

All reported issues were addressed across 132 files

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

Fix all with cubic | Re-trigger cubic</sub>