feat(runtime): complete host-local inference lifecycle

[ericksoa]

Summary

This draft completes provider-neutral lifecycle handling for durable host-local inference receipts across OpenClaw, Hermes, and DCode with Ollama, NIM, and vLLM. Production activation remains dormant.

Changes

• Persist and re-prove host-local inference ownership through snapshot manifests and backups.
• Re-prove exact authority before backup publication and across restore mutation fences.
• Carry exact receipts through rebuild, self-restore, and cross-sandbox clone flows.
• Retire replaced managed runtimes safely while retaining a runtime referenced by another sandbox.
• Add provider-neutral destroy behavior with exact-ID, identity-checked, idempotent Podman cleanup for managed NIM/vLLM.
• Retain externally owned Ollama services and preserve registry ownership when teardown requires retry.\n- Prove retry convergence after confirmed deletion and require idempotent provider teardown.\n- Remove an incomplete snapshot clone if its live inference authority cannot be re-proved.\n- Require every lifecycle re-proof to perform fresh provider-native authority inspection and network health probing.
• Exercise an MXC-style provider through the generic contract without Podman-specific central switches.
• Inherit exact managed route retry and explicit ownership-clear coverage from feat(runtime): route host-local inference #8066 and feat(runtime): persist host-local inference ownership #8069.
• Keep the implementation dormant; native Podman support is not advertised by this slice.

Quality Gates

• Tests added or updated for changed behavior
• Docs not applicable: production activation is unchanged
• Sensitive paths changed
• Exact-head advisor, CI, and protected E2E refresh completed

Documentation Writer Review

• Documentation writer subagent reviewed the completed changes
• Result: pending exact-head refresh
• Evidence: review requested for exact head 75cd0af36c2be335f24603c7e612237ed0f10c58.
• Agent: Codex Desktop

Verification

• Signed commits and DCO
• Exact slice plus inherited contract suite: 13 files, 178 tests passed
• npm run build:cli
• npm run typecheck:cli
• npm --prefix nemoclaw run typecheck
• npm run checks:repository
• Test-conditional growth scan and normal commit/push hooks passed
• Exact-head GitHub qualification is running

Stack

• Depends on feat(runtime): persist host-local inference ownership #8069.
• Tracks epic [Epic] Support native Podman with buildless managed onboarding #7744.
• Exact base: 990a5f5634d3df19c7d60b96e0e4b8c438936f20.
• Exact head: 75cd0af36c2be335f24603c7e612237ed0f10c58.
• Stable base-relative patch ID: 5a966b4d9e939ea378f428cd0664ce59392cc441.

---

Signed-off-by: Aaron Erickson aerickson@nvidia.com

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 20266ef5-1ee8-480b-8caf-ac5a2c214a25

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-code-quality]

Code Coverage Overview

Languages: TypeScript

TypeScript / code-coverage/plugin

The overall coverage in commit 75cd0af in the feat/host-local-infe... branch remains at 96%, unchanged from commit f93b27d in the feat/host-local-infe... branch.

Show a code coverage summary of the most impacted files.

File	feat/host-local-infe... f93b27d	feat/host-local-infe... 75cd0af	+/-
nemoclaw/src/se...ntial-filter.ts	100%	0%	-100%
nemoclaw/src/sh...er-boundary.cts	95%	0%	-95%
nemoclaw/src/se...ot-sanitizer.ts	95%	0%	-95%
nemoclaw/src/co...test-fixture.ts	81%	0%	-81%
nemoclaw/src/co...ration-state.ts	93%	94%	+1%

TypeScript / code-coverage/cli

The overall coverage in commit 75cd0af in the feat/host-local-infe... branch remains at 81%, unchanged from commit f93b27d in the feat/host-local-infe... branch.

Show a code coverage summary of the most impacted files.

File	feat/host-local-infe... f93b27d	feat/host-local-infe... 75cd0af	+/-
src/lib/onboard...trap/adapter.ts	0%	63%	+63%
src/lib/onboard...strap/docker.ts	0%	64%	+64%
src/lib/onboard...ne-lifecycle.ts	0%	83%	+83%
src/lib/onboard...cker-journal.ts	0%	83%	+83%
src/lib/onboard...al-inference.ts	0%	83%	+83%
src/lib/onboard...trap-journal.ts	0%	86%	+86%
src/lib/actions...ne-providers.ts	0%	86%	+86%
src/lib/onboard...-replacement.ts	0%	87%	+87%
src/lib/onboard...-transaction.ts	0%	88%	+88%
src/lib/onboard...test-fixture.ts	0%	96%	+96%

---

_{Updated August 02, 2026 20:10 UTC}

[github-actions]

PR Review Advisor — No blocking findings reported

Advisor assessment: No blocking advisor findings reported
Next action: No advisor follow-up needed.
Findings: 0 blockers · 0 warnings · 0 suggestions

Model lanes

• GPT-5.6 Terra (primary): Completed · high confidence · 0 blockers · 0 warnings · 0 suggestions
• Nemotron 3 Ultra (second opinion): Completed · high confidence · 0 blockers · 0 warnings · 0 suggestions
• Model comparison: normalized findings match; normalized terminology decisions differ; normalized E2E selections differ; severity counts match.

3 terminology differences from the second opinion

Advisory only. These are normalized differences from the primary terminology receipt.

• retire at src/lib/onboard/runtime-provider/host-local-inference-lifecycle.ts:149: selected only by the second-opinion lane as justified.
• reprove at src/lib/onboard/runtime-provider/host-local-inference-lifecycle.ts:54: selected only by the second-opinion lane as justified.
• host-local inference authority at src/lib/onboard/runtime-provider/host-local-inference-lifecycle.ts:62: selected only by the second-opinion lane as define.

Second-opinion terminology and E2E selections are advisory. They do not change the primary assessment or E2E / PR Gate.

4 semantic terminology decisions

Terminology decisions are advisory. They affect the assessment only when a separate finding identifies concrete semantic impact.

• define — host-local inference authority at src/lib/onboard/runtime-provider/host-local-inference-lifecycle.ts:73: Define this term where it first appears in user-visible or contributor-facing explanatory text if that surface is added; retain the exact phrase for this lifecycle concept.
• define — provider snapshot authority at src/lib/actions/sandbox/snapshot.ts:1136: Use this term only for the combined provider-backed snapshot boundary, and retain managed snapshot authority when only managed workload/runtime authority is meant.
• define — exact provider ownership at src/lib/actions/sandbox/snapshot.ts:1372: Name the retained registry ownership record when this phrase appears in user-visible or contributor-facing explanatory text.
• established — mutation fence at src/lib/actions/sandbox/snapshot.ts:1429: Retain mutation fence for the immediate pre-mutation validation point.

E2E guidance

Advisory only. E2E / PR Gate selects and runs jobs independently.

Recommended E2E: inference-routing, network-policy, onboard-repair, onboard-resume, rebuild-openclaw, state-backup-restore, cloud-onboard

2 optional E2E recommendations

• snapshot-commands
• sandbox-operations

Workflow run details

This automated review informs maintainers. Warnings and suggestions do not require a response. A maintainer decides whether to merge.