Opus 4.6 and Codex 5.3 updates by rblalock · Pull Request #874 · agentuity/sdk

rblalock · 2026-02-05T19:28:46Z

opencode modules --refresh

To get the latest.

Also, removed the delegate tool since we should be using the built in Open Code task tool automatically now.

Summary by CodeRabbit

Updates
- Agents upgraded to newer model versions (OpenAI GPT-5.3-codex, Claude Opus 4-6).
- Legacy delegate tool removed and related public exports cleaned up.
- Monitor agent now allows an additional previously-excluded tool.
Bug Fixes
- Background task notifications deduplicated to avoid repeated status updates.
Documentation
- Docs updated to describe new task-based and background delegation approach.
Tests
- Tests adjusted to reflect updated model version strings.

coderabbitai · 2026-02-05T19:29:08Z

📝 Walkthrough

Walkthrough

Bumped default models for several agents; added per-task notification deduplication via a new notifiedStatuses field; removed the delegate tool and its public exports and plugin hook; updated docs, examples, and tests to reflect model and tooling changes.

Changes

Cohort / File(s)	Summary
Agent Model Updates `packages/opencode/src/agents/architect.ts`, `packages/opencode/src/agents/builder.ts`, `packages/opencode/src/agents/lead.ts`	Updated agent `defaultModel` strings: Architect `openai/gpt-5.2-codex` → `openai/gpt-5.3-codex`; Builder & Lead `anthropic/claude-opus-4-5-20251101` → `anthropic/claude-opus-4-6`.
Background Task Deduplication `packages/opencode/src/background/manager.ts`, `packages/opencode/src/background/types.ts`	Added per-task notification deduplication: new optional `notifiedStatuses?: Set<BackgroundTaskStatus>` on `BackgroundTask` and manager logic to skip duplicate status notifications and handle terminal-state edge cases.
Delegate Tool Removal `packages/opencode/src/tools/delegate.ts`, `packages/opencode/src/tools/index.ts`, `packages/opencode/src/plugin/plugin.ts`	Removed the `delegate` tool module, its exported types/schemas (`DelegateArgsSchema`, `DelegateArgs`), AGENT_MENTIONS helper, and the `agentuity_coder_delegate` plugin hook; cleaned up related exports and plugin tooling surface.
Docs & Delegation Pattern Changes `packages/opencode/AGENTS.md`, `packages/opencode/src/agents/reasoner.ts`	Removed delegate.ts reference and JSON-based delegation description; documentation and examples updated to show Task and background-task mechanisms instead; `reasonerAgent.tools.exclude` no longer contains `"task"`.
Agent Examples & Monitor `packages/opencode/src/agents/memory.ts`, `packages/opencode/src/agents/monitor.ts`	Updated JSON example model identifiers in memory examples to `claude-opus-4-6`; removed `agentuity_coder_delegate` from `monitorAgent.tools.exclude` so it is not excluded.
Tests & Config Samples `packages/opencode/test/agents.test.ts`, `packages/opencode/test/config.test.ts`, `packages/opencode/src/config/loader.ts`	Adjusted tests and sample config comment strings to new model versions (`openai/gpt-5.3-codex`, `anthropic/claude-opus-4-6`) and updated assertions/expectations accordingly.

🚥 Pre-merge checks | ✅ 1

✅ Passed checks (1 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

📜 Recent review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between a597dae and 37dec7f.

📒 Files selected for processing (3)

packages/opencode/src/agents/reasoner.ts
packages/opencode/src/plugin/plugin.ts
packages/opencode/test/agents.test.ts

💤 Files with no reviewable changes (1)

packages/opencode/src/plugin/plugin.ts

🚧 Files skipped from review as they are similar to previous changes (1)

packages/opencode/src/agents/reasoner.ts

🧰 Additional context used

📓 Path-based instructions (4)

**/*.{ts,tsx}

📄 CodeRabbit inference engine (AGENTS.md)

**/*.{ts,tsx}: Use Prettier formatter with tabs (width 3), single quotes, and semicolons for TypeScript files
Use TypeScript strict mode with ESNext target and bundler moduleResolution
Use StructuredError from @agentuity/core for error handling

Files:

packages/opencode/test/agents.test.ts

packages/*/test/**/*.{ts,tsx}

📄 CodeRabbit inference engine (AGENTS.md)

packages/*/test/**/*.{ts,tsx}: Place test files in test/ folder (never in src/ or __tests__/)
Import from ../src/ in test files to reference source code
Use @agentuity/test-utils for mocks and test helpers

Files:

packages/opencode/test/agents.test.ts

packages/opencode/**/*.{ts,tsx}

📄 CodeRabbit inference engine (packages/opencode/AGENTS.md)

Use Zod for schema validation

Files:

packages/opencode/test/agents.test.ts

packages/opencode/test/**/*.{ts,tsx}

📄 CodeRabbit inference engine (packages/opencode/AGENTS.md)

When running tests, prefer using a subagent (Task tool) to avoid context bloat from test output

Files:

packages/opencode/test/agents.test.ts

🧠 Learnings (1)

📚 Learning: 2025-12-21T00:31:41.858Z

Learnt from: jhaynie
Repo: agentuity/sdk PR: 274
File: packages/cli/src/cmd/build/vite/server-bundler.ts:12-41
Timestamp: 2025-12-21T00:31:41.858Z
Learning: In Bun runtime, BuildMessage and ResolveMessage are global types and are not exported from the bun module. Do not import { BuildMessage } from 'bun' or similar; these types are available globally and should be used without import. This applies to all TypeScript files that target the Bun runtime within the repository.

Applied to files:

packages/opencode/test/agents.test.ts

🧬 Code graph analysis (1)

packages/opencode/test/agents.test.ts (1)

packages/opencode/src/agents/index.ts (1)

agents (20-35)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (3)

GitHub Check: SDK Integration Test Suite
GitHub Check: Sandbox CLI Tests
GitHub Check: Cloud Deployment Tests

🔇 Additional comments (2)

packages/opencode/test/agents.test.ts (2)

105-111: Model expectation update looks correct.

Aligns the Architect default model assertion with the new gpt-5.3 codex value.

148-153: Tool exclusion assertion looks correct.

The Reasoner test now correctly allows the task tool.

_{✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.}

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

github-actions · 2026-02-05T19:32:15Z

📦 Canary Packages Published

version: 1.0.1-37dec7f

Packages

Package	Version	URL
`@agentuity/react`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-react-1.0.1-37dec7f.tgz
`@agentuity/cli`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-cli-1.0.1-37dec7f.tgz
`@agentuity/workbench`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-workbench-1.0.1-37dec7f.tgz
`@agentuity/server`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-server-1.0.1-37dec7f.tgz
`@agentuity/auth`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-auth-1.0.1-37dec7f.tgz
`@agentuity/postgres`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-postgres-1.0.1-37dec7f.tgz
`@agentuity/drizzle`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-drizzle-1.0.1-37dec7f.tgz
`@agentuity/frontend`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-frontend-1.0.1-37dec7f.tgz
`@agentuity/runtime`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-runtime-1.0.1-37dec7f.tgz
`@agentuity/opencode`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-opencode-1.0.1-37dec7f.tgz
`@agentuity/schema`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-schema-1.0.1-37dec7f.tgz
`@agentuity/core`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-core-1.0.1-37dec7f.tgz
`@agentuity/evals`	`1.0.1-37dec7f`	https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-evals-1.0.1-37dec7f.tgz

Install

Add to your package.json:

{
  "dependencies": {
    "@agentuity/react": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-react-1.0.1-37dec7f.tgz",
    "@agentuity/cli": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-cli-1.0.1-37dec7f.tgz",
    "@agentuity/workbench": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-workbench-1.0.1-37dec7f.tgz",
    "@agentuity/server": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-server-1.0.1-37dec7f.tgz",
    "@agentuity/auth": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-auth-1.0.1-37dec7f.tgz",
    "@agentuity/postgres": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-postgres-1.0.1-37dec7f.tgz",
    "@agentuity/drizzle": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-drizzle-1.0.1-37dec7f.tgz",
    "@agentuity/frontend": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-frontend-1.0.1-37dec7f.tgz",
    "@agentuity/runtime": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-runtime-1.0.1-37dec7f.tgz",
    "@agentuity/opencode": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-opencode-1.0.1-37dec7f.tgz",
    "@agentuity/schema": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-schema-1.0.1-37dec7f.tgz",
    "@agentuity/core": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-core-1.0.1-37dec7f.tgz",
    "@agentuity/evals": "https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-evals-1.0.1-37dec7f.tgz"
  }
}

Or install directly:

bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-react-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-cli-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-workbench-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-server-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-auth-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-postgres-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-drizzle-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-frontend-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-runtime-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-opencode-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-schema-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-core-1.0.1-37dec7f.tgz
bun add https://agentuity-sdk-objects.t3.storageapi.dev/npm/1.0.1-37dec7f/agentuity-evals-1.0.1-37dec7f.tgz

coderabbitai

Actionable comments posted: 2

🤖 Fix all issues with AI agents

In `@packages/opencode/AGENTS.md`:
- Around line 132-135: Update the AGENTS.md text to remove the nonexistent "Task
tool" blocking delegation mechanism and instead describe only the available
parallel mechanism: `agentuity_background_task` (with
`agentuity_background_output` and `agentuity_background_cancel`), ensuring the
doc language reflects that delegation is done via a background task/session for
concurrent work and that no blocking sequential "task" tool exists; also verify
the tools export list in code matches the documentation (i.e.,
`agentuity_background_task`, `agentuity_background_output`,
`agentuity_background_cancel` are the described APIs).

In `@packages/opencode/src/agents/reasoner.ts`:
- Around line 154-160: The prompt in agents/reasoner.ts instructs the agent to
"Use the Task tool" but the agent's configuration currently blocks that tool via
tools.exclude, causing a mismatch; either remove "task" from the agent's
tools.exclude list (so the Task tool is available to the Reasoner agent) or
change the prompt text to remove/replace the Task-specific instruction (e.g.,
instruct to use an available tool such as Memory or a generic tool) and make the
same change for the other occurrence referenced in the file; update the
configuration or the prompt string in the Reasoner module so the instruction and
available tools align.

📜 Review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 56de09c and a597dae.

📒 Files selected for processing (6)

packages/opencode/AGENTS.md
packages/opencode/src/agents/monitor.ts
packages/opencode/src/agents/reasoner.ts
packages/opencode/src/plugin/plugin.ts
packages/opencode/src/tools/delegate.ts
packages/opencode/src/tools/index.ts

💤 Files with no reviewable changes (4)

packages/opencode/src/agents/monitor.ts
packages/opencode/src/tools/index.ts
packages/opencode/src/plugin/plugin.ts
packages/opencode/src/tools/delegate.ts

🧰 Additional context used

📓 Path-based instructions (3)

**/*.{ts,tsx}

📄 CodeRabbit inference engine (AGENTS.md)

**/*.{ts,tsx}: Use Prettier formatter with tabs (width 3), single quotes, and semicolons for TypeScript files
Use TypeScript strict mode with ESNext target and bundler moduleResolution
Use StructuredError from @agentuity/core for error handling

Files:

packages/opencode/src/agents/reasoner.ts

packages/opencode/**/*.{ts,tsx}

📄 CodeRabbit inference engine (packages/opencode/AGENTS.md)

Use Zod for schema validation

Files:

packages/opencode/src/agents/reasoner.ts

packages/opencode/src/agents/*.ts

📄 CodeRabbit inference engine (packages/opencode/AGENTS.md)

Agent system prompts must be embedded in agent definition files

Files:

packages/opencode/src/agents/reasoner.ts

🧠 Learnings (1)

📚 Learning: 2025-12-21T00:31:41.858Z

Learnt from: jhaynie
Repo: agentuity/sdk PR: 274
File: packages/cli/src/cmd/build/vite/server-bundler.ts:12-41
Timestamp: 2025-12-21T00:31:41.858Z
Learning: In Bun runtime, BuildMessage and ResolveMessage are global types and are not exported from the bun module. Do not import { BuildMessage } from 'bun' or similar; these types are available globally and should be used without import. This applies to all TypeScript files that target the Bun runtime within the repository.

Applied to files:

packages/opencode/src/agents/reasoner.ts

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (12)

GitHub Check: Framework Integration Tests (TanStack & Next.js)
GitHub Check: Storage CLI Tests
GitHub Check: Playwright E2E Smoke Test
GitHub Check: Sandbox CLI Tests
GitHub Check: Standalone Agent Test
GitHub Check: Queue CLI Tests
GitHub Check: SDK Integration Test Suite
GitHub Check: Package Installation & Usage Test
GitHub Check: Cloud Deployment Tests
GitHub Check: Queue SDK Tests
GitHub Check: Template Integration Tests
GitHub Check: Build

_{✏️ Tip: You can disable this entire section by setting review_details to false in your review settings.}

packages/opencode/AGENTS.md

packages/opencode/src/agents/reasoner.ts

model updates

3519ba3

rblalock requested a review from jhaynie February 5, 2026 19:28

Merge branch 'main' into opus46-codex53

56de09c

Merge branch 'main' into opus46-codex53

cb5b0d1

rblalock enabled auto-merge (squash) February 5, 2026 19:34

jhaynie approved these changes Feb 5, 2026

View reviewed changes

remove delegate

a597dae

coderabbitai bot reviewed Feb 5, 2026

View reviewed changes

packages/opencode/AGENTS.md Show resolved Hide resolved

packages/opencode/src/agents/reasoner.ts Show resolved Hide resolved

rblalock added 2 commits February 5, 2026 15:25

fix task tool

7d89b29

remove agent mentinos delegate tool

37dec7f

rblalock merged commit 34e64b2 into main Feb 5, 2026
15 checks passed

rblalock deleted the opus46-codex53 branch February 5, 2026 20:32

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Opus 4.6 and Codex 5.3 updates#874

Opus 4.6 and Codex 5.3 updates#874
rblalock merged 6 commits intomainfrom
opus46-codex53

rblalock commented Feb 5, 2026 •

edited by coderabbitai bot

Loading

Uh oh!

coderabbitai bot commented Feb 5, 2026 •

edited

Loading

Walkthrough

Changes

Uh oh!

github-actions bot commented Feb 5, 2026 •

edited

Loading

Uh oh!

coderabbitai bot left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

rblalock commented Feb 5, 2026 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Uh oh!

coderabbitai bot commented Feb 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Uh oh!

github-actions bot commented Feb 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📦 Canary Packages Published

Uh oh!

coderabbitai bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

rblalock commented Feb 5, 2026 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Feb 5, 2026 •

edited

Loading

github-actions bot commented Feb 5, 2026 •

edited

Loading