OpenClaw and Hermes already have a first-day post-release reliability queue
The next useful thing to watch is not another headline feature; it is whether the new tagged releases settle down under real operators. OpenClaw reports now include production event-loop delays, cron jobs skipping instead of trying cloud fallbacks when a local primary model is down, Gemini subagents hanging at stream-ready with zero tokens, session-store recovery after crash/OOM writes, Codex route preservation, and OpenRouter model-id normalization. Hermes reports include automatic TUI heap dumps growing to tens of GiB, skill frontmatter names diverging from directory names, MiniMax OAuth expiry parsing, Telegram media uploads hitting a 20s write-timeout path, and Feishu table rendering regressions in v0.13.0.
The releases are still important, but the first few hours after a broad agent-runtime release often reveal the operator-facing edge cases: fallback decisions, long-running sessions, provider routing, disk pressure, and channel formatting. Those are exactly the paths that decide whether an agent can keep working without babysitting.
- OpenClaw issue #79327 reports repeated production event-loop delays exceeding 2s with uncaught errors
- OpenClaw issue #79329 reports cron model preflight skipping an entire run when a local primary provider is unreachable, despite configured cloud fallbacks
- OpenClaw issue #79333 reports Google Gemini subagents hanging after stream-ready with 0 tokens while direct infer works
- OpenClaw PR #79319 adds best-effort session-store recovery from backup and temp artifacts after crash/OOM during atomic writes
- OpenClaw PR #79316 preserves explicit openai-codex provider routes, and PR #79313 normalizes OpenRouter model ids
- Hermes issue #21767 reports automatic TUI heap snapshots growing ~/.hermes/heapdumps to tens of GiB; PR #21780 makes auto heap dumps opt-in
- Hermes issue #21782 shows skill create/edit can diverge frontmatter name from directory name
- Hermes issue #21779 and PR #21786 cover MiniMax OAuth treating a unix-ms expiry timestamp as TTL seconds
- Hermes PR #21777 wires Telegram media_write_timeout so large media uploads are not stuck on PTB defaults
- Hermes issue #21778 reports Feishu/Lark markdown tables rendering as raw source in v0.13.0
- Most fixes are still PR-stage, not released tags
- Some reports are provider- or channel-specific, so teams that do not use Gemini, MiniMax, Telegram, Feishu, Codex, or OpenRouter may only need to monitor them
- The volume of fresh issues is high; publishable value here is the cluster pattern, not any one isolated bug