Guide
v2026.5.16-beta.4 is easiest to understand through failure. The official record describes a provider, model, voice, or context change; the sections below spell out what to save, what to interrupt, and what must remain available afterward.
This candidate is still pre-release. Treat provider, channel, and migration behavior as unconfirmed until your own bounded test passes. Treat the version as one input to the decision. Local credentials, enabled plugins, channels, and host support still need their own evidence.
Channel
Pre-Release
Primary signals
Security/audit; Agents/subagents; Providers/media
Publication date
2026-05-17
Failure and recovery path
Security/audit
Evidence from the release. add security.audit.suppressions for intentionally accepted audit findings, keeping suppressed matches out of the active summary while preserving them in JSON output with an active suppression notice.
Operator reading. Before widening access for Security/audit, replay a narrow allow and deny case. A readable denial and a revoked test credential are part of the acceptance record.
Agents/subagents
Evidence from the release. label delegated task and subagent completion handoffs as ready for parent review, and tell requester agents to review/verify results before calling them done.
Operator reading. Keep Agents/subagents inside a test workspace until you can inspect its ledger, next run, failure state, and cancellation behavior from the operator surface.
Providers/media
Evidence from the release. add fal and OpenRouter music-generation providers for the shared music_generate tool, including fal MiniMax/ACE/Stable Audio endpoints and OpenRouter Lyria audio output.
Operator reading. The safe reading of Providers/media is operational: prove the model can answer the intended task, then force a bounded failure and verify that no side effect is repeated.
Agents/subagents
Evidence from the release. route group/channel subagent completions through message-tool-only handoffs when required and keep active-requester wake failures from dropping completion delivery. Fixes #82803.
Operator reading. The operator question around Agents/subagents is where the reply lands after a restart. Record the conversation identifiers before testing and compare them after the channel is healthy again.
The limits of the recovery promise
Memory-core
scan persisted memory source sessions on startup, comparing on-disk transcripts against the index and marking only missing/newer/resized files dirty for incremental sync. Fixes #82341.
Measure Memory-core after a fresh session and after a reconnect. The record should say what was retained, what was intentionally absent, and which identity requested the recall.
Run the smallest useful rehearsal
- Write down the baseline for Security/audit, including the installed version, host, provider, and workspace.
- Confirm what can be restored, then take a verified backup before applying the candidate package.
- Exercise the affected channel with synthetic text and note the thread, sender, reaction, or attachment id that comes back.
- Restart or reconnect the Gateway, then check for one intended reply rather than a duplicate, missing, or misrouted message.
- Keep the experiment bounded; an unexpected result belongs in the release record before any wider rollout.
openclaw --version
openclaw gateway status
openclaw security auditLeave a usable maintenance note
Save the command, output, source link, and environment used for this check. Future operators should be able to tell whether a difference belongs to the release or to local provider, channel, or platform state.
Fix details to carry forward
Agents/subagents. route group/channel subagent completions through message-tool-only handoffs when required and keep active-requester wake failures from dropping completion delivery. Fixes #82803. This is the edge case to include in a regression check, especially when the release is pre-release.
Memory-core. scan persisted memory source sessions on startup, comparing on-disk transcripts against the index and marking only missing/newer/resized files dirty for incremental sync. Fixes #82341. Treat the fix as a boundary condition and retain the source wording beside your local result.
Keep the full v2026.5.16-beta.4 notes with your test result. Read the official release index for neighboring versions and the model provider reference for configuration limits.
Reference Trail
Sources and further reading
- full v2026.5.16-beta.4 notesgithub.com
- official release indexgithub.com
- model provider referencedocs.openclaw.ai