diff --git a/docs/inference/declarative-agents-manifest.mdx b/docs/inference/declarative-agents-manifest.mdx index 4e65ed3633c..4f22d9d7b3e 100644 --- a/docs/inference/declarative-agents-manifest.mdx +++ b/docs/inference/declarative-agents-manifest.mdx @@ -169,6 +169,7 @@ Workspaces under `/sandbox/.openclaw/workspace-` are preserved across rebuil The runtime startup script provisions them on first boot rather than baking their contents. For ad-hoc per-agent edits inside an existing sandbox without a rebuild, use the in-sandbox CLI, `nemoclaw agents add|delete|list`. +To reconcile the whole roster against a manifest in one step, use `nemoclaw agents apply -f `, described in [Apply to an Existing Sandbox](#apply-to-an-existing-sandbox). Use the manifest for fixed, checked-in layouts. Use the CLI passthrough for interactive work. diff --git a/docs/inference/switch-providers.mdx b/docs/inference/switch-providers.mdx index abd40c393d4..4935a211083 100644 --- a/docs/inference/switch-providers.mdx +++ b/docs/inference/switch-providers.mdx @@ -35,7 +35,7 @@ Run `$$nemoclaw onboard` to register the provider, then retry the switch. Pass the target provider and model together. -Pass `--sandbox ` when you do not want to use the default sandbox. +Target a non-default sandbox by name: the `shields` commands take it as a positional argument, and `inference set` takes `--sandbox `. ```bash $$nemoclaw shields down @@ -61,10 +61,10 @@ This flag does not bypass shared-gateway compatibility checks. -## Recreate a Deep Agents Sandbox - +## Recreate a Deep Agents Sandbox + Deep Agents uses fresh sandbox recreation for provider changes. The recreation validates the provider and rewrites `/sandbox/.deepagents/config.toml` with the OpenShell route.