Weekly Developer Report (2026-08-04 to 2026-08-10)

Claude Code Updates

- **Claude Code 2.1.225 Released**
- [CLI Changes Overview](https://x.com/ClaudeCodeLog/status/2085899665665339399)
- Highlights:
- 14 CLI changes, including an edit tool for precise string replacements.
- New usage warnings for gateway spend caps.

- **Cross-session Messaging**
- [Announcement](https://x.com/claudedevs/status/2085817074816070014)
- Users can now communicate between sessions to improve workflow continuity.

- **New Default Permission Mode**
- Starting August 14, **auto mode** will be the default for Pro, Max, and Team users.
- Significantly improves safety by catching 89% of dangerous commands.

- **Multiple Releases**:
- Version [2.1.224](https://x.com/ClaudeCodeLog/status/2085580408490451419) and [2.1.223](https://x.com/ClaudeCodeLog/status/2085170539719483394) introduced CLI changes for local execution and GitHub organization policies respectively.
- [2.1.226](https://x.com/ClaudeCodeLog/status/2085924447572082805) focuses on bug fixes and reliability improvements.

- **Managed Agents Updates**
- [New budgeting feature](https://x.com/claudedevs/status/2085853169930957158) allows users to pause sessions once budget limits are reached.

GitHub Copilot Updates

- **New Code Review Effort Levels**
- [Release Info](https://github.blog/changelog/2026-08-07-copilot-code-review-effort-levels-are-generally-available)
- Introduces **Lite** and **Balanced** effort levels to improve code review alignment with project complexity.

- **Usage Metrics API Update**
- Analytics now include agent app activity to track tools like Claude and Codex better.
- [More on the updates](https://github.blog/changelog/2026-08-07-copilot-usage-metrics-api-adds-agent-app-activity).

- **Secret Scanning Coverage**
- [Coverage Updates](https://github.blog/changelog/2026-08-07-secret-scanning-coverage-updates) now include more secrets and added a partner, **Lovable Labs**.

- **Code Quality Enhancements**
- Automatic code coverage generation feature is now available for faster setup in workflows.
- [Details here](https://github.blog/changelog/2026-08-04-code-coverage-automatic-enablement-in-code-quality-settings).

Codex Updates

- **Codex Security Review Launch**
- [Research Preview](https://x.com/OpenAIDevs/status/2085482310636560830) aims to analyze GitHub pull requests for security issues using repository context.

- **Model Access Expansion**
- Improved GPT-5.6 Sol for accuracy; expanded free access to GPT-5.6 Luna.
- [Improvements detailed](https://openai.com/index/improving-gpt-5-6-sol-in-chatgpt).

- **Voice Feature in Codex**
- [New Feature](https://x.com/OpenAIDevs/status/2085068107958411518) allows verbal input, enhancing session initiation for coding.

Additional Highlights

- **Agent Plugins**
- [Cross-platform Plugin Support](https://x.com/OpenAIDevs/status/2085398373511918022) introduced, enabling a unified plugin experience across various clients.

- **Copper and New UI Components**
- Copper now enables file attachments, enhancing workflow.
- [New Questionnaire Component](https://x.com/shadcn/status/2085083695401078912) simplifies multi-step question flows.

- **Security Fixes**
- Recent updates ensure sensitive data is redacted during operations.
- [Fix for secrets](https://github.com/mattpocock/skills/commit/efce423018fc6468a3239621f1c1bcaacc723801).

For full detail and policies regarding each tool, consult the provided links.