How the Claude Code team does code review
How do you review code at the volume agents produce it?
Why it matters: A new product capability you may be able to use right away, without code changes.
Anthropic publishes no official RSS; tracked via a community mirror of its official pages + official GitHub Releases.
118 updates · page 1 of 8
How do you review code at the volume agents produce it?
Why it matters: A new product capability you may be able to use right away, without code changes.
How one business development representative clears his inbound queue and drafts bilingual customer emails with Claude Managed Agents and Claude Tag.
Why it matters: A new product capability you may be able to use right away, without code changes.
Claude Agent SDK (Python) v0.2.168 updates the bundled Claude CLI to version 2.1.301.
Why it matters: Python users of the Claude Agent SDK can install 0.2.168 to pick up Claude CLI 2.1.301.
Claude Code v2.1.301 adds a usage-limit wait cap for unattended retry, OpenTelemetry guidance in the claude-api skill, and a vim-mode percent motion to jump to a matching bracket.
Why it matters: Claude Code users can stop unattended retries that would wait too long and jump between brackets in vim mode.
Anthropic published an update on independent evaluators.
Why it matters: Readers tracking Anthropic's evaluation process can see the latest note on independent evaluators.
Claude Code Review is now generally available, aimed at catching more bugs before they reach production.
Why it matters: Teams using Claude can review code for bugs before those bugs ship to production.
Claude Agent SDK (Python) v0.2.167 updates the bundled Claude CLI to version 2.1.300.
Why it matters: Python SDK users who install claude-agent-sdk 0.2.167 get the bundled Claude CLI at version 2.1.300.
Claude Code v2.1.300 adds virtualText prompt decorations that never enter the draft, splits OpenTelemetry cache writes into cache_creation_5m_tokens and cache_creation_1h_tokens, and adds $.ui.view for showing an agent's transcript.
Why it matters: Claude Code users on v2.1.300 can decorate the prompt without submitting that text and can split cache-write telemetry by lifetime.
Claude Agent SDK (Python) v0.2.166 adds typed hook output classes and input fields, plus a SystemInitData TypedDict for the init system message payload.
Why it matters: Hook authors on the Python SDK can use typed outputs and SystemInitData instead of untyped dicts in v0.2.166.
Claude Code v2.1.299 adds CLAUDE_CODE_DISABLE_FILE_MENTIONS so @ mentions pass paths rather than file contents, and idle compaction for a 200K+ token conversation left idle for about 54 minutes.
Why it matters: Terminal users can stop file mentions from injecting contents, and idle chats compact before the one-hour prompt cache expires.
Anthropic has opened a Singapore office and signed an MOU with the Government of Singapore.
Why it matters: People working with Anthropic in Singapore now have a local office and a signed government MOU.
A new update covers preparing for the 2026 US midterms.
Why it matters: Readers following election-related work can see this note on preparing for the 2026 US midterms.
The Claude Platform is announcing OpenTelemetry traces from Claude Managed Agents, the Messages API, and Claude SDKs.
Why it matters: Developers can collect OpenTelemetry traces from Claude Managed Agents, the Messages API, and Claude SDKs.
Anthropic migrated to an in-house version control system in five months after merges per day grew more than tenfold and source control became its biggest development bottleneck.
Why it matters: Engineering teams facing the same bottleneck can read how Anthropic redesigned source control.
Uber's CISO, Philip Martin, discusses monitoring and observability for AI agents and how Uber's open-source tooling works alongside Enterprise Frontier Safeguards.
Why it matters: Security leads can read how Uber watches AI agents and pairs that open-source tooling with Enterprise Frontier Safeguards.