clawstat.us
Plum Comet Labs

Should you update OpenClaw v2026.11.2? — Update with care

Verdict: Update with care (medium confidence), assessed 2026-11-10.

Evidence: 1 credible blocking issue confirmed for this version, none widespread.

Best version to run today: v2026.10.10 — 7 days in the field, 1 credible known issue, none widespread; last verdict Update with care. If you are on a newer version that works for you, stay put. The latest, v2026.11.2, came out 2 days ago — too new to call settled.

Update with care: the core is functional, but Slack users face a gateway freeze, Codex delegation is fragile, and model-session bugs remain open

Why this verdict

v2026.11.2 is a large release (166 commits, 113 PRs) whose main confirmed blocker is #196249: with Slack streaming set to "off", a long reasoning stream reprocesses accumulated text on every delta and can pin the Gateway main thread at 100% CPU until SIGKILL — Slack, WhatsApp, cron and SQLite admission all time out while it runs. This is the one credible blocking issue the evidence stands behind, and it is contained: it requires Slack DM sessions with streaming explicitly disabled plus a long-reasoning model, and it has drawn no community engagement (0 👍, 1 comment) despite ~2.9M weekly npm downloads. A fix PR is linked but open, and the newest stable does not include it; a beta line (v2026.11.5-beta.1, with beta.2 already referenced in #196674) is brewing, but nothing has shipped yet.

Known issues (60+)

  • #196249 (critical) [Bug]: Gateway freezes during a long reasoning stream in a Slack DM when Slack streaming is off
  • #196199 (high) [Bug]: Regression since 2026.10.10: native Codex helper completions fail to resume the parent, pausing goals and locking users out of their assistant
  • #196142 (high) [Bug]: OpenRouter catalog rows built right after Gateway start say reasoning:false, so every thinking level clamps to off and reasoning-mandatory models reject the request
  • #195265 (high) [Bug]: First config reload after a gateway start marks every agent's model runtime stale (auth snapshot captured in two shapes)
  • #195269 (high) [Bug]: Gateway startup defers every agent when the first agent databases in target order have no clean-close seal
  • #196248 (high) Claude Fable 5.2 (`claude-fable-5-2`) unsupported in 2026.11.2: no catalog row, direct-path Claude Code identity below Anthropic's floor, prefix-check contract not applied
  • #196505 (high) memory index aborts with "worker task capacity reached" instead of retrying: a ~864 MiB corpus never completes, while the same build indexes ~205 MiB in 1.6 min
  • #196544 (high) [Bug]: Configured model without maxTokens silently gets an 8192 output cap; truncation mid tool call surfaces only as 'incomplete tool call'

ClawStat.us gives an independent, evidence-based answer to whether you should update OpenClaw — for every release. Is the latest OpenClaw update safe, or should you wait for the next one? Each verdict is built from post-release bug reports and a two-model review, refreshed every few hours.

Machine-readable: JSON API · llms.txt · full markdown

Independent, unofficial project — not affiliated with or endorsed by OpenClaw or its maintainers. Verdicts are generated automatically and may be wrong; confirm against the linked issues before updating. Data is drawn from the public github.com/openclaw/openclaw issue tracker and releases via the GitHub API.