Repository Issue Activity (beta)

#8649 [DGX Spark][Hermes] current LKG: gateway restart quarantines agent; rebuild can leave registry/live OpenShell divergent

#8648 Windows checkout converts the starter prompt to CRLF and breaks repository checks

#8306 [Ubuntu][Upgrade] nemoclaw update --fresh reinstalls the fixed "lkg" tag with no downgrade guard, silently downgrading a newer install (v0.0.102 -> v0.0.97) while printing "already up to date"

#8438 build-typecheck broken on main: duplicate `readinessComparisonMatches` in serving/resolver.ts (TS2393)

#8549 [Ubuntu 26.04][Agent&Skills] contributor-plan-issue skill: "plan issue {url}" trigger does not produce the required structured plan report

#7744 [Epic] Support native Podman with buildless managed onboarding

#8600 Define OpenShell v0.0.101 qualification receipts and gates

#8590 [Epic] Qualify and pin OpenShell v0.0.101

#8599 Review OpenShell v0.0.101 migration and security contracts

#8598 Record OpenShell v0.0.101 baselines and release trust

#2544 Plugin config (plugins.entries.<id>.config in openclaw.json) is policy-locked at runtime — every config change requires multi-minute image rebuild

#5795 [Linux][Onboard] nemoclaw onboard silently degrades to best_effort Landlock mode on kernel < 5.13 instead of aborting — sandbox created with reduced isolation

#5801 rebuild/upgrade: atomic build-verify-swap before deleting old sandbox (health-before-delete)

#6005 [All Platforms][Onboard][GitHub Issue #6005] nemoclaw onboard wizard has no back-navigation — users cannot correct a previous step without Ctrl+C

#6492 refactor(state): make registry and gateway lifecycle updates atomic

#6660 Stabilize full-e2e cold-onboard performance budget

#6691 Run speculative compatibility tests against candidate OpenShell, OpenClaw, Hermes, and DCode versions

#6850 [Ubuntu][Inference] managed gateway health probe sends max_tokens to GPT-5 endpoints — probe fails HTTP 400 and gates agent traffic (incomplete fix of #6642)

#7127 DGX Spark managed vLLM: MTP cold-start NV_ERR_NO_MEMORY, negative memory estimate, and long-context agent instability

#7140 ci(e2e): improve reliability and observability of long-running runners

#7144 perf(e2e): reduce rebuild-hermes runtime with pinned fixtures and reusable build cache

#7145 ci(e2e): move heavy Docker lanes to larger GitHub-hosted runners

#7146 ci(e2e): classify hosted-runner resource pressure and infrastructure loss

#7206 ci(e2e): reconcile successful same-diff reruns with immutable PR gate checks

#7288 chore(security): track fix-forward hardening for historical OpenClaw revisions