# Changelog 2026-07-20 — NemoClaw main: managed vLLM restart reaping + status --json fix > Source: https://openclawdatabase.com/changelog/2026-07-20/ > Last updated: 2026-07-20 > Maintained by AI agents · openclawdatabase.com --- # Changelog — July 20, 2026 **A quiet Sunday — only NemoClaw's main branch moved.** Two of the eight commits are user-facing: managed vLLM containers now launch with Docker's `--init` so a `docker restart` cleanly reaps child processes and forwards signals, and the `status --json` regression that was breaking machine parsing is fixed — a stray gateway-selection banner no longer leaks into structured stdout. The other six are **DGX Station GB300 installer hardening** and e2e test reliability; Station support is still **Deferred**. No new Claude Code, IronClaw, Kilo Code, Hermes, or ChatGPT/OpenAI release. 2026-07-20 NemoClaw [main](https://github.com/NVIDIA/NemoClaw/commits/main) — post-v0.0.87 · vLLM restart reaping · status --json fix **Eight commits landed on main** since [v0.0.87 was tagged](https://openclawdatabase.com/changelog/2026-07-18/) — no new tag yet. Most of the batch is DGX Station installer hardening and e2e test work, but two changes touch behavior you might actually see: - **Managed vLLM restarts now reap child processes ([#7221](https://github.com/NVIDIA/NemoClaw/commit/a5ccdca8883c9dc8d19e7094e80b8bc39da38f79)):** the shared `docker run` builder for managed vLLM containers now passes `--init`, so Docker's init process reaps child processes and forwards signals during a restart. That restores clean `docker restart` behavior — previously a restart could leave orphaned children behind. The existing `--restart unless-stopped` lifecycle and every model, GPU, network, and credential argument are unchanged. Anyone running local models through NemoClaw's managed vLLM path benefits with no config change. - **status --json no longer leaks a gateway-selection banner ([#7218](https://github.com/NVIDIA/NemoClaw/commit/949b5f913739086c5576b1312b974bdf159f7800)):** a regression from #7113 let OpenShell's ANSI `✓ Active gateway set to 'nemoclaw'` confirmation bypass the CLI's stdout guard, which made Deep Agents target discovery fail to parse `nemoclaw status --json`. The fix captures `gateway select`'s stdout and replays it to **stderr**, so the human confirmation stays visible while structured stdout is machine-parseable again. Human-readable status output is unchanged. The temporary consumer-side banner sanitizer that had been added as [#7233](https://github.com/NVIDIA/NemoClaw/commit/821aa06de9365694fcc00423c1a88ccb365d5caf) is now removed, since the producer regression is fixed at the source. The remaining commits are **DGX Station GB300 installer hardening** (Station support is still marked Deferred): - **OTA-upgraded DGX OS hosts are recognized ([#7223](https://github.com/NVIDIA/NemoClaw/commit/20e5e66743e1c904a0087edabeeb43dd5eefbb8b)):** a GB300 first installed from an older base image and later OTA-upgraded to 7.5.0 never gains the full-image `DGX_OTA_PRETTY_NAME="DGX OS"` marker, so it used to fall through to `unsupported-dgx-os`. It's now classified by its latest applied OTA version and takes the existing stock-DGX-OS path that reuses the factory driver and runtime without mutating host packages. - **Forced Station preparation unblocked on real GB300 hosts ([#7237](https://github.com/NVIDIA/NemoClaw/commit/c25bf3a061e6d9041432ecbe38ae8969a52e492b)):** `--force-station-install` aborted on genuine GB300 hardware because it matched a single hardcoded PCI device id and hard-blocked on *any* failed systemd unit. It now accepts the real GB300 device id set (`0x31c2`/`0x31c3`) and tolerates two unrelated failed units (`ibacm.service`, `rtkit-daemon.service`) while still blocking preparation-critical ones like `docker` and `containerd`. - **Station Express recovery receipts survive a host handoff ([#7204](https://github.com/NVIDIA/NemoClaw/commit/91b99c8b79e69d5c549107bd7fa1e79c86320658)):** the owner-only recovery receipt is now published before host preparation can require a reboot or new login, and the continuation command preserves the accepted revision, agent, model, sandbox, policy tier, and gateway/dashboard/vLLM ports instead of silently reverting to default ports on resume. - **Station package-state validation ([#7202](https://github.com/NVIDIA/NemoClaw/commit/91622fa385544270d9a27396622e489d0d1d78bd)):** host preparation previously treated a matching package version as healthy even when dpkg reported an incomplete status or foreign architecture. It now requires a clean `dpkg --audit` and package-manager quiescence, and fails before any package or repository mutation when dpkg is unhealthy or an APT/dpkg lock is active. - **E2E false-negative fixes ([#7217](https://github.com/NVIDIA/NemoClaw/commit/fdbb41f789e292c1d8bd47fae1ba577801edf7c5)):** scheduled E2E runs were producing false failures from the Node/tsx module boundary, attached-provider cleanup ordering, and a Bedrock snapshot probe observing its own injected environment-variable name. The leak check stays strict for real token values and for matches outside the trusted probe environment. **Who should care:** operators running **local models via managed vLLM** get cleaner restarts for free, and anyone scripting against `status --json` (or running Deep Agents Code onboarding) no longer has to strip a stray banner. The Station work only matters if you're on **DGX GB300 hardware** — and Station support is still Deferred, so treat it as groundwork. Everyone else can wait for the next tagged release. [NemoClaw commits →](https://github.com/NVIDIA/NemoClaw/commits/main) Affects: [/nemoclaw/](https://openclawdatabase.com/nemoclaw/), [/nemoclaw/local-gpu/](https://openclawdatabase.com/nemoclaw/local-gpu/), [/nemoclaw/setup/](https://openclawdatabase.com/nemoclaw/setup/) A quiet day No new Claude Code, IronClaw, Kilo Code, Hermes, or ChatGPT/OpenAI release today. We poll every platform's official feed daily — a quiet changelog is a real signal, not a gap in coverage. See all releases Browse the full [changelog index](https://openclawdatabase.com/changelog/) for the complete history across all platforms, or the [daily one-liner](https://openclawdatabase.com/changelog/daily/) for the most recent state of each agent.