Skip to content

Commit 214f1c4

Browse files
authored
Merge branch 'main' into security/strip-credentials-and-blueprint-digest
2 parents c405b78 + 4cfc3b3 commit 214f1c4

105 files changed

Lines changed: 8130 additions & 2128 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎.agents/skills/docs/nemoclaw-configure-inference/SKILL.md‎

Lines changed: 0 additions & 59 deletions
This file was deleted.

‎.agents/skills/docs/nemoclaw-reference/references/inference-profiles.md‎

Lines changed: 0 additions & 53 deletions
This file was deleted.
Lines changed: 85 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,85 @@
1+
---
2+
name: nemoclaw-configure-inference
3+
description: Changes the active inference model without restarting the sandbox. Use when change inference runtime, inference routing, openclaw, openshell, switch nemoclaw inference model, switch nemoclaw inference models.
4+
---
5+
6+
# Nemoclaw Configure Inference
7+
8+
Change the active inference model without restarting the sandbox.
9+
10+
## Prerequisites
11+
12+
- A running NemoClaw sandbox.
13+
- The OpenShell CLI on your `PATH`.
14+
15+
Change the active inference model while the sandbox is running.
16+
No restart is required.
17+
18+
## Step 1: Switch to a Different Model
19+
20+
Switching happens through the OpenShell inference route.
21+
Use the provider and model that match the upstream you want to use.
22+
23+
### NVIDIA Endpoints
24+
25+
```console
26+
$ openshell inference set --provider nvidia-prod --model nvidia/nemotron-3-super-120b-a12b
27+
```
28+
29+
### OpenAI
30+
31+
```console
32+
$ openshell inference set --provider openai-api --model gpt-5.4
33+
```
34+
35+
### Anthropic
36+
37+
```console
38+
$ openshell inference set --provider anthropic-prod --model claude-sonnet-4-6
39+
```
40+
41+
### Google Gemini
42+
43+
```console
44+
$ openshell inference set --provider gemini-api --model gemini-2.5-flash
45+
```
46+
47+
### Compatible Endpoints
48+
49+
If you onboarded a custom compatible endpoint, switch models with the provider created for that endpoint:
50+
51+
```console
52+
$ openshell inference set --provider compatible-endpoint --model <model-name>
53+
```
54+
55+
```console
56+
$ openshell inference set --provider compatible-anthropic-endpoint --model <model-name>
57+
```
58+
59+
If the provider itself needs to change, rerun `nemoclaw onboard`.
60+
61+
## Step 2: Verify the Active Model
62+
63+
Run the status command to confirm the change:
64+
65+
```console
66+
$ nemoclaw <name> status
67+
```
68+
69+
Add the `--json` flag for machine-readable output:
70+
71+
```console
72+
$ nemoclaw <name> status --json
73+
```
74+
75+
The output includes the active provider, model, and endpoint.
76+
77+
## Step 3: Notes
78+
79+
- The host keeps provider credentials.
80+
- The sandbox continues to use `inference.local`.
81+
- Runtime switching changes the OpenShell route. It does not rewrite your stored credentials.
82+
83+
## Related Skills
84+
85+
- `nemoclaw-reference` — Inference Profiles for full profile configuration details
File renamed without changes.

.agents/skills/docs/nemoclaw-get-started/SKILL.md renamed to .agents/skills/nemoclaw-get-started/SKILL.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -70,7 +70,7 @@ When the install completes, a summary confirms the running environment:
7070
```text
7171
──────────────────────────────────────────────────
7272
Sandbox my-assistant (Landlock + seccomp + netns)
73-
Model nvidia/nemotron-3-super-120b-a12b (NVIDIA Endpoint API)
73+
Model nvidia/nemotron-3-super-120b-a12b (NVIDIA Endpoints)
7474
──────────────────────────────────────────────────
7575
Run: nemoclaw my-assistant connect
7676
Status: nemoclaw my-assistant status
File renamed without changes.
File renamed without changes.
File renamed without changes.

.agents/skills/docs/nemoclaw-overview/references/how-it-works.md renamed to .agents/skills/nemoclaw-overview/references/how-it-works.md

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -93,7 +93,8 @@ After the sandbox starts, the agent runs inside it with all network, filesystem,
9393

9494
Inference requests from the agent never leave the sandbox directly.
9595
OpenShell intercepts every inference call and routes it to the configured provider.
96-
NemoClaw routes inference to NVIDIA Endpoints, specifically Nemotron 3 Super 120B through [build.nvidia.com](https://build.nvidia.com). You can switch models at runtime without restarting the sandbox.
96+
During onboarding, NemoClaw validates the selected provider and model, configures the OpenShell route, and bakes the matching model reference into the sandbox image.
97+
The sandbox then talks to `inference.local`, while the host owns the actual provider credential and upstream endpoint.
9798

9899
## Network and Filesystem Policy
99100

.agents/skills/docs/nemoclaw-overview/references/overview.md renamed to .agents/skills/nemoclaw-overview/references/overview.md

File renamed without changes.

0 commit comments

Comments
 (0)