Claude Code 2.1.295, Codex 0.162 and Copilot Stacked PRs Go GA
Yesterday's news was about models and sandboxes. Today's is about control: hooks that fail closed in Claude Code, managed Git worktrees in Codex, and stacked pull requests that finally make large agent-written changes reviewable on GitHub. Details come from the Claude Code changelog, the Codex changelog and the GitHub changelog.
Claude Code 2.1.295: Hooks That Fail Closed
- onFailure: "block" for command and HTTP hooks means a hook that errors or times out now blocks the action instead of letting it through. For security guardrails this is the safe default you were missing.
- Terminal status (OSC 7501): Program Status Protocol support lets terminals show whether Claude is working, waiting or done. The Claude apps gateway also returns a request-id header so telemetry can be matched to the gateway audit log.
- MCP changes: tool descriptions loaded through tool search are now cut at 16,384 characters instead of 2,048, subagents preload at most 32 skills, and claude.ai connectors negotiate MCP protocol version 2026-07-28 by default. Many MCP reconnect and pagination bugs are fixed.
- Fixes: requests failing on [1m] models when a gateway or cloud provider refuses the context-1m beta, and background-work turns dropping earlier output in claude -p.
OpenAI Codex: CLI 0.161, 0.162 and Sol Ultrafast
Codex CLI 0.162.0 (October 8, see the release list) adds managed Git worktree tools for trusted local projects, task pinning in the agent Command Center, capability settings for custom model providers, and JavaScript helpers with ranked tool search in Code Mode. Fixes include CRLF preservation in apply_patch, Linux sandbox and Windows file-access issues, honoring server Retry-After advice, and a signed PowerShell installer for Windows.
Codex CLI 0.161.0 (October 7) made GPT-6.1 Sol the default model in the bundled and Amazon Bedrock catalogs, added multi-agent V2 and Ultra reasoning on Bedrock, and introduced /mcp login <name> for signing in to MCP servers.
GPT-6.1 Sol Ultrafast speeds up token generation in Codex and ChatGPT Work. It is available on Pro $500 and eligible Enterprise and Edu plans, is off by default for Enterprise until a workspace owner enables it, and supports inference residency in the US and Europe (EEA plus Switzerland).
GitHub Copilot: Stacked PRs and Code Review API
Stacked pull requests are GA on all github.com plans, per the announcement. You can split a large change into small, independently reviewable PRs that merge together. "Rebase stack" keeps approvals on unchanged code, a stack enters the merge queue as one merge group, and a stack is retargeted instead of closed if its base branch is deleted. Shift+J and Shift+K navigate the stack, the pull_request webhook has a stacked action, and the gh stack extension supports worktrees. Auto-merge for a group is rolling out over the next few weeks, and GitHub Enterprise Server support is planned.
Copilot code review gets an API. The October 2 changelog says reviews can be requested through the REST and GraphQL APIs with a per-request effort setting, and Balanced is now the default level (explicit Lite selections were kept). It is available on Pro, Pro+, Max, Business and Enterprise.
Also today: maintainers can now count draft pull requests toward PR limits, a useful brake on low-quality, agent-generated contributions.
OpenCode
The OpenCode changelog shows no release since v1.18.35 (October 6), which added JSON and Markdown formats for agent-readable stats and image support in xAI tool results. A quiet day for OpenCode.
What to Do Today
- Update Claude Code to 2.1.295 and set onFailure: "block" on any hook that guards destructive actions.
- Upgrade Codex CLI to 0.162.0 and try managed worktrees for parallel agent tasks.
- Break large agent-made changes into stacked PRs, and wire Copilot code review into CI through the new API.
Sources
Comments
Share your thoughts and join the conversation
