Back to all articles

OpenClaw 2026.5.7 Release Notes: ChatGPT Instant Alias, Cron Status JSON, and Channel CLI Changes

OpenClaw v2026.5.7 was released on May 7, 2026. According to the official GitHub release, this is a fix-heavy update focused on publishing reliability, CLI clarity, session correctness, and authorization hardening.

If you have been following the rapid release cycle, this update continues the cleanup work from OpenClaw 2026.5.6 and OpenClaw 2026.5.5, but with a broader spread across plugins, cron, channels, memory, and session handling.

What Changed in OpenClaw 2026.5.7

The most visible change for many users is new support for openai/chat-latest as an explicit direct API-key model override. The release notes say this lets users try the moving ChatGPT Instant API alias without changing the stable default model. For teams testing model behavior before switching production defaults, that is a practical addition.

Another important change is in the cron CLI. The official release notes state that openclaw cron list --json and openclaw cron show --json now include a computed top-level status field. The official cron documentation shows the expected values as disabled, running, ok, error, skipped, or idle. That matters for anyone building dashboards or automation around scheduled jobs because external tooling no longer needs to derive job status itself.

The channels CLI also changed. The release notes say openclaw channels list is now channel-only, while openclaw channels list --all shows bundled and catalog channels. The official channels documentation now reflects that split and notes that configured accounts expose installed, configured, and enabled state. In practice, this should make channel inventory and troubleshooting more predictable.

Security and Authorization Fixes

Several May 7 fixes are clearly about reducing unsafe behavior rather than adding new features.

The release notes say native command handlers now honor owner enforcement. They also state that Active Memory global toggles now require admin scope. In the same release, inline skill tool dispatch for auto-reply was gated through before-tool-call authorization hooks. Taken together, these changes tighten who can trigger sensitive operations and who can change memory-related behavior.

There is also a credential-handling fix for Tavily. Per the official release notes, dedicated tavily_search and tavily_extract credentials are now resolved from the active runtime config snapshot so SecretRef-backed keys do not reach tools unresolved.

Session and Delivery Reliability

OpenClaw 2026.5.7 includes several fixes aimed at long-running agents and multi-channel delivery.

The release notes state that cached skill snapshots are now cleared during /new and sessions.reset, so long-lived channel sessions rebuild the visible skill list after changes. Another fix invalidates cached assembled context views when source history shrinks or context assembly fails, which is meant to prevent stale history from being reused.

On the delivery side, the official release notes mention:

  • Discord channel targets prefixed like discord:channel:<id> are now parsed as channel sends instead of being misrouted as legacy DM targets.
  • Empty outbound adapter results now report deliverySucceeded=false instead of appearing successful.
  • Isolated cron runs now fail announce delivery before model execution when delivery.channel=last has no previous route, which avoids spending model tokens before hitting a permanent delivery error.

If you run scheduled tasks or route agents across multiple chat surfaces, those are meaningful operational fixes.

Why This Release Matters

This is not a headline feature release like OpenClaw 2026.5.4, which introduced the realtime Gemini voice bridge for Google Meet dial-in joins. Instead, OpenClaw 2026.5.7 is the kind of release that makes production setups less ambiguous and less fragile.

For operators, the most valuable changes are probably:

  • better machine-readable cron status output
  • clearer channel inventory commands
  • stronger authorization around native commands and memory toggles
  • more accurate delivery and session cache behavior

If your priority is stable automation, those are high-value maintenance changes.

Official Sources

By CompareClaw TeamUpdated May 2026