[rsliter]

Outcome

The inference guides now distinguish new hosted custom provider attachments from Personal-tier and retained shared routes for OpenClaw, Hermes, Deep Agents Code, and Pi.

Reason

Published routing summaries, operational references, and the generated support matrix still described hosted custom selections with the old route matrix.

Related issues

Fixes #13341

Changes

• Correct routing, readback, policy, monitoring, security, and credential guidance across the affected agent variants.
• Update the support matrix source and generated reference table to match the current route rules.
• Explain direct endpoint and Bedrock adapter boundaries, including sandbox credential placeholders.
• Clarify that inference set needs an existing native attachment to switch a custom endpoint onto the native route; retained shared-route sandboxes stay shared.
• Preserve Pi's activation warning and the remaining live qualification gaps.

Verification

• npm run docs passed on commit ade13d3d83d1c1cdb7586c08073c3e4d4b83f6ef: agent variants, route checks, and Fern validation completed with zero errors.
• python3 scripts/generate-platform-docs.py --check and git diff --check passed.
• Repository pre-commit and pre-push hooks passed, including the secret scan, publication validation, and CLI TypeScript check.
• Independent documentation review of the exact commit passed with no remaining findings.
• This change updates documentation and support metadata; it does not change runtime behavior.

---

Signed-off-by: Rebecca Sliter 571084+rsliter@users.noreply.github.com

Summary by CodeRabbit

• Documentation
  • Clarified inference routing for hosted, local, and host-local provider selections.
  • Documented that eligible hosted selections outside the Personal tier and qualified local selections can use sandbox-attached providers.
  • Clarified when Personal-tier selections, unattached endpoints, legacy shared routes, and Model Router use shared routing.
  • Added guidance on endpoint-bound credentials, host-managed secrets, and Bedrock adapter handling.
  • Updated setup, switching, status, troubleshooting, deployment, security, and platform-support guidance.
  • Documented qualification and activation requirements for selected providers and runtimes.

[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]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Repository: NVIDIA/NemoClaw/.coderabbit.yaml
• Review profile: CHILL
• Plan: Enterprise
• Run ID: 1cc68a4b-3866-4e72-9d1f-a65e2de2f0be

📥 Commits

Reviewing files that changed from the base of the PR and between 6669f70 and ade13d3.

📒 Files selected for processing: 22

• ci/platform-matrix.json
• docs/about/how-it-works.mdx
• docs/about/overview.mdx
• docs/deployment/deploy-to-headless-server.mdx
• docs/get-started/quickstart-pi.mdx
• docs/inference/custom-endpoint-security.mdx
• docs/inference/how-inference-routing-works.mdx
• docs/inference/set-up-anthropic-compatible-endpoint.mdx
• docs/inference/set-up-openai-compatible-endpoint.mdx
• docs/inference/switch-providers.mdx
• docs/inference/understand-provider-validation.mdx
• docs/inference/use-shared-gateway-routes.mdx
• docs/inference/view-active-inference-route.mdx
• docs/monitoring/monitor-sandbox-activity.mdx
• docs/reference/architecture.mdx
• docs/reference/commands.mdx
• docs/reference/enterprise-readiness.mdx
• docs/reference/network-policies.mdx
• docs/reference/pi-support.mdx
• docs/reference/platform-support.mdx
• docs/reference/troubleshooting.mdx
• docs/security/best-practices.mdx

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 8 remain after this review.

📝 Walkthrough

The pull request updates inference-routing documentation across onboarding, command references, support matrices, and security guidance. It distinguishes attached-provider routes from inference.local and documents credential handling and qualification limits.

Changes: Inference routing documentation

Layer / File(s)	Summary
Routing and credential contract docs/about/*, docs/inference/custom-endpoint-security.mdx, docs/inference/how-inference-routing-works.mdx	The general routing and credential guidance distinguishes hosted, local, Personal-tier, retained, and Bedrock cases.
Endpoint setup and provider selection docs/get-started/quickstart-pi.mdx, docs/inference/set-up-*-compatible-endpoint.mdx, docs/inference/switch-providers.mdx, docs/inference/understand-provider-validation.mdx, docs/inference/use-shared-gateway-routes.mdx, docs/inference/view-active-inference-route.mdx	Setup and provider-selection guidance describes attached and shared routes, credential references, and qualification conditions.
Operational checks and commands docs/deployment/deploy-to-headless-server.mdx, docs/reference/commands.mdx, docs/monitoring/monitor-sandbox-activity.mdx, docs/reference/troubleshooting.mdx	Readiness, command, monitoring, and troubleshooting documentation adopts the revised native-versus-shared route distinctions.
Support, security, and platform references ci/platform-matrix.json, docs/reference/architecture.mdx, docs/reference/enterprise-readiness.mdx, docs/reference/network-policies.mdx, docs/reference/pi-support.mdx, docs/reference/platform-support.mdx, docs/security/best-practices.mdx	Reference tables and guidance describe routing categories, credential boundaries, and provider qualification limits.

Priority: ⬇️ Low

Estimated code review effort: 3 (Moderate) | ~25 minutes

Change: Other · Severity of issue fixed: Low

Possibly related issues

• Stability ahead of OpenShell 0.1.3 upgrade #13338 — The epic tracks the documentation corrections for hosted routing covered by this pull request.

Possibly related PRs

• NVIDIA/NemoClaw#13363: It validates native routing for Deep Agents Code hosted compatible endpoints and updates related routing claims.
• NVIDIA/NemoClaw#12889: It implements the sandbox-bound provider behavior and credential contract described in this documentation.
• NVIDIA/NemoClaw#13292: It changes Hermes native-local attachment and credential-reference handling documented here.

Suggested reviewers: prekshivyas, sandl99

Merge Risk: ⚪ Minimal · up to ade13

The checked guidance preserves host-managed credentials and directs Local NVIDIA NIM users to the shared-route diagnostics. No actionable merge-blocking issue was established.

Pre-merge checks: 5

Check name	Status	Explanation
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.
Title check		The title clearly and concisely describes the primary change: correcting documentation for hosted custom inference routing.
Linked Issues check		Issue #13341 requires corrected hosted-native versus retained shared routing for Hermes, Deep Agents Code, and Pi, with consistent examples, credentials, host custody, placeholders, API limits, Pi sta…
Out of Scope Changes check		The changes stay within the documentation and support-metadata scope of issue #13341. The broader edits update related routing, readback, policy, monitoring, security, credential, command, troubleshoo…
Docstring Coverage		No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…

Full details: Linked Issues check

Explanation

Issue #13341 requires corrected hosted-native versus retained shared routing for Hermes, Deep Agents Code, and Pi, with consistent examples, credentials, host custody, placeholders, API limits, Pi status, durable endpoint binding, and the three live qualification gaps. The reviewed pages document endpoint-bound attachments outside Personal, inference.local for Personal or retained routes, host-local Bedrock adapters, scoped credential references, Pi activation limits, and no fallback after native-selection failure. docs/inference/how-inference-routing-works.mdx preserves the binding rule that removing an ambient allowlist entry does not revoke a verified binding. The Hermes Anthropic-to-OpenAI gap, fresh hosted custom OpenAI gap, and Pi credential-rewrite request gap remain explicit. The reported docs, generator, link, and review checks address the required documentation validation.

Full details: Out of Scope Changes check

Explanation

The changes stay within the documentation and support-metadata scope of issue #13341. The broader edits update related routing, readback, policy, monitoring, security, credential, command, troubleshooting, and generated support guidance so the affected rules remain consistent across public references. No runtime behavior change or unrelated product feature is shown.

Full details: Docstring Coverage

Explanation

No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0 files. (22 skipped: 22 unsupported.)

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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

[github-actions]

🌿 Preview your docs: https://nvidia-preview-pr-13609.docs.buildwithfern.com/nemoclaw

[github-code-quality]

Code coverage overview

Languages: TypeScript

TypeScript / code-coverage/plugin

main 63002cd	codex/13341-routing-... ade13d3	+/-
96%	97%	+1%

View coverage of the 10 most impacted files.

File	main 63002cd	codex/13341-routing-... ade13d3	+/-
nemoclaw/src/onboard/config.ts	98%	96%	-2%
nemoclaw/src/index.ts	94%	93%	-1%
nemoclaw/src/bl...print/runner.ts	95%	95%	0%
nemoclaw/src/bl...ime-identity.ts	97%	97%	0%
nemoclaw/src/bl...t-management.ts	100%	100%	0%
nemoclaw/src/co.../config-show.ts	100%	100%	0%
nemoclaw/src/commands/slash.ts	100%	100%	0%
nemoclaw/src/ru...time-context.ts	100%	100%	0%
nemoclaw/src/sh...er-boundary.cts	100%	100%	0%
nemoclaw/src/on...native-route.ts	0%	100%	+100%

TypeScript / code-coverage/cli

main 63002cd	codex/13341-routing-... ade13d3	+/-
84%	86%	+2%

View coverage of the 10 most impacted files.

File	main 63002cd	codex/13341-routing-... ade13d3	+/-
src/lib/onboard.ts	62%	50%	-12%
src/lib/actions...ess-recovery.ts	65%	81%	+16%
src/lib/state/l...diness-lease.ts	66%	83%	+17%
src/lib/inferen...ollama/proxy.ts	41%	61%	+20%
src/lib/onboard.../application.ts	55%	84%	+29%
src/lib/domain/...y/dimensions.ts	0%	71%	+71%
src/lib/inferen...er/lifecycle.ts	0%	73%	+73%
src/lib/actions...ry/operation.ts	0%	76%	+76%
src/lib/actions...try/snapshot.ts	0%	78%	+78%
src/lib/inferen...ovider/index.ts	0%	88%	+88%

ℹ️ Calculations represent line coverage. Results may change as calculations are updated.

---

_{Updated October 21, 2026 22:15 UTC}

[github-actions]

PR Review Advisor finished for commit ade13d3. Include the Advisor findings in the complete PR feedback collection. Verify and group valid findings before repair.

Request review only when Require no Advisor blockers is green.

All previous runs

[rsliter]

E2E verification for commit ade13d3d83d1c1cdb7586c08073c3e4d4b83f6ef

The PR Review Advisor selected the deterministic cloud-onboard floor for ci/platform-matrix.json. All nine specialists recommended no additional E2E selectors.

Selector	Result	Evidence
jobs=cloud-onboard, targets= empty, inference_mode=mock	Passed: Cloud onboard (docker) and Relevant E2E	Run, cloud-onboard job, aggregate

The trusted dispatch receipt binds this run to PR #13609, candidate commit ade13d3d83d1c1cdb7586c08073c3e4d4b83f6ef, and base/workflow commit 6669f7047fc05e3a9897f87411a1fe9e2810d45f.