Official agent skill

Debug Inference

by NVIDIA in NVIDIA/OpenShell

Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM.

OfficialApache-2.0Auto-check passedAI & LLM Engineering

Install Debug Inference

skills CLI
$ npx skills add NVIDIA/OpenShell --skill debug-inference -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install NVIDIA/OpenShell debug-inference --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/NVIDIA/OpenShell.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/debug-inference .claude/skills/debug-inference && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
debug-inference
GitHub stars
15k
Token cost
~1.9k tokens
SKILL.md length
848 words
Files
1
Skills in repo
23
Repo updated
First seen
Licence
Apache-2.0

At a glance

Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM.

  • Works in 5 steps: Confirm Gateway and Sandbox Context → Inspect the Provider and Its Profile → Confirm Attachment → …
  • Provider attachment
  • SKILL.md covers Diagnostic Workflow, Host-Local Inference Checklist and Reporting
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Debug Inference is an agent skill from NVIDIA/OpenShell, published by the product's own GitHub organization. Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM. Use for provider attachment, endpoint policy, credential substitution, topology, and migration from the removed managed inference endpoint. Trigger keywords - debug inference, managed inference endpoint, local inference, ollama, lm studio, vllm, sglang, trtllm, NIM, inference failing, model server unreachable, credentialendpointmismatch…

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in AI & LLM Engineering, covering LLM inference and serving. It works with Ollama, SGLang and vLLM. The repository describes itself as: OpenShell is the safe, private runtime for autonomous AI agents. The licence is Apache-2.0.

When your agent uses it

  • Provider attachment
  • Endpoint policy
  • Credential substitution
  • Migration from the removed managed inference endpoint

Example prompts

  • “/debug-inference”

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. Confirm Gateway and Sandbox Context
  2. Inspect the Provider and Its Profile
  3. Confirm Attachment
  4. Verify Native Client Configuration
  5. Interpret Common Failures

What it can do on your machine

Read from SKILL.md and the folder at commit 834b79a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • docs.nvidia.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Debug Inference loads about 1.9k tokens when it runs. Until then it costs about 137 tokens; SKILL.md has 848 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~137
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from NVIDIA/OpenShell at commit 834b79a, republished under its Apache-2.0 licence (© NVIDIA). 848 words, ~1,881 tokens.

Download SKILL.mdSave it as .claude/skills/debug-inference/SKILL.md (or your agent's skills folder).
name
debug-inference
description
Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM. Use for provider attachment, endpoint policy, credential substitution, topology, and migration from the removed managed inference endpoint. Trigger keywords - debug inference, managed inference endpoint, local inference, ollama, lm studio, vllm, sglang, trtllm, NIM, inference failing, model server unreachable, credential_endpoint_mismatch, host.openshell.internal.

Debug Inference

Diagnose inference as ordinary provider-authorized network traffic. OpenShell no longer supplies a managed inference route, rewrites request shapes, or selects a model. The application calls the provider's native endpoint and owns its base URL, model, request format, and timeout.

Use installed openshell --help output as the authority for command syntax. Refer to the published provider management guide and provider profile guide for current behavior.

Diagnostic Workflow

1. Confirm Gateway and Sandbox Context
bash
openshell status
openshell gateway info
openshell sandbox get <sandbox>

For a host-local model server, host.openshell.internal identifies the machine running the gateway. It does not identify the operator's laptop when the gateway is remote. A server listening only on 127.0.0.1 may also be unreachable from a container; bind it to an address reachable from the gateway runtime.

2. Inspect the Provider and Its Profile
bash
openshell provider get <provider>
openshell profile export <profile-id> -o yaml

Check that the profile:

  • Names the exact endpoint host, port, and protocol the client calls.
  • Allows the client binary.
  • Declares the credential key and intended authentication style.
  • Uses narrow HTTP rules when the provider should expose only part of an API.

For a custom or self-hosted OpenAI-compatible endpoint, import an endpoint-bearing profile. A base URL stored only in provider configuration does not authorize a new endpoint.

bash
openshell profile lint -f ./provider-profile.yaml
openshell profile import -f ./provider-profile.yaml
openshell provider create --name <provider> --type <profile-id>

Add the required --credential KEY or --credential KEY=VALUE arguments shown by the profile. Never broaden endpoint policy merely to silence a credential binding error.

3. Confirm Attachment
bash
openshell sandbox provider list <sandbox>
openshell sandbox provider attach <sandbox> <provider> --wait --timeout 30

Save the change's receipt_id and use openshell sandbox provider status <sandbox> <provider> --receipt <receipt-id> --wait --timeout 30 to check when it takes effect. Success confirms that the sandbox applied the credentials, policy, and environment for new processes. If the result is pending, failed, withheld, or superseded, inspect its reason before launching the client.

Launch the client after the attachment wait succeeds so it receives the updated environment:

bash
openshell sandbox exec <sandbox> -- <client-command>

After updating an ordinary static provider, wait for that change and launch a new client. An existing process keeps its revision-scoped reference; readiness does not make the old reference resolve the replacement value. Diagnose managed-refresh credentials according to their own lifecycle.

Keep credentials and issued references out of diagnostic output. Acknowledged detach revokes future credential resolution and removes the reference from future process environments. Requests already forwarded may still finish:

bash
openshell sandbox provider detach <sandbox> <provider> --wait --timeout 30
4. Verify Native Client Configuration

The application must use the real upstream contract:

  • Native provider base URL, not the retired managed virtual endpoint.
  • Real model ID, not a placeholder that OpenShell used to rewrite.
  • Native OpenAI, Anthropic, Vertex, or other provider request shape.
  • Application-owned timeout and retry settings.
  • The credential environment variable declared by the attached profile.

Probe the exact endpoint from a newly launched sandbox process. Start with a non-secret discovery endpoint when the provider offers one, then send a minimal inference request using the provider's documented API shape.

Show full SKILL.md (400 more words)Show less
5. Interpret Common Failures
SymptomLikely causeFix
credential_placeholder_in_request_bodyA body reference is invalid/revoked, or classification metadata is unavailableCheck the controlled denial reason; remove the reference from conversation history or restore provider access. Do not enable body credential rewriting or bypass flags to send tool output. Unknown literals and valid issued placeholders pass unchanged, including the model provider’s own placeholder. Header resolution does not enable body rewriting.
A retired managed endpoint fails DNS resolutionClient still uses the removed managed endpointConfigure the provider's native base URL and attach an endpoint-bearing provider profile
Direct request is deniedMissing attachment, endpoint policy, HTTP rule, or binary authorizationInspect the attached provider profile and sandbox effective policy
credential_endpoint_mismatchCredential profile does not authorize the request recipientCorrect the host/port/path or import a narrowly scoped profile for the intended endpoint
request_authority_mismatchHTTP authority differs from the CONNECT destinationUse the same host and effective port in both authorities
Credential variable is absentProvider was not attached when this process launched, or profiles collide on a keyAttach the provider and launch a new process; resolve duplicate keys explicitly
Upstream rejects the model or bodyClient relied on removed model/request rewritingConfigure the real model and provider-native request format in the application
127.0.0.1 works on the host but not in the sandboxLoopback refers to different runtimeUse host.openshell.internal or another gateway-reachable endpoint and profile
Host-local request times outServer bind address, gateway topology, or host firewall blocks container-to-host trafficVerify the listener and permit only the required gateway network path and port

Host-Local Inference Checklist

For Ollama, LM Studio, vLLM, SGLang, TRT-LLM, and local NIM deployments:

  1. Verify the engine from the gateway host.
  2. Verify it listens on an address reachable from the gateway runtime.
  3. Import a custom profile naming host.openshell.internal and the actual port.
  4. Restrict the profile to the intended binaries and API paths.
  5. Create and attach the provider.
  6. Configure the application's base URL, model, and timeout.
  7. Probe the native endpoint from a newly launched sandbox process.

Reporting

Report:

  1. The active gateway and whether topology contributes to the failure.
  2. The provider, profile, attachment, endpoint, and client binary involved.
  3. The exact failed host, port, path, and request authority without secrets.
  4. Whether the client still relies on removed managed-routing behavior.
  5. The narrowest profile, attachment, or application configuration change that resolves the problem.

© NVIDIA, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/debug-inference of NVIDIA/OpenShell.

Open the folder on GitHubat commit 834b79a

Compare with similar skills

Debug Inference next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Debug Inference compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Debug Inference this skillNVIDIA/OpenShell15k—~1.9kAutomated safety check: PassApache-2.0
Dev BumpNetis/heron102—~983Automated safety check: PassApache-2.0
Source Command Dev Docs SyncNetis/heron102—~1.4kAutomated safety check: PassApache-2.0
Agentsop LLM Engine Selectionagentsope/SkillAlchemy457—~6.1kAutomated safety check: PassMIT
SageMaker Serving Image Selectionhuggingface/skills11k1 repos~4.6kAutomated safety check: PassApache-2.0
Aider DelegateamElnagdy/delegate-skills2.3k2 repos~3kAutomated safety check: PassMIT

Similar skills

  • Dev Bump

    Netis/heron

    Bump Heron version via the VERSION-file SSOT. An agent skill from Netis/heron.

    102 GitHub stars~983 tokensUpdated 3 days ago
    AI & LLM EngineeringAuto-check passed
  • Sync documentation with project state using ICAV workflow. An agent skill from Netis/heron.

    102 GitHub stars~1.4k tokensUpdated 3 days ago
    AI & LLM EngineeringAuto-check passed
  • Agentsop LLM Engine Selection

    agentsope/SkillAlchemy

    Cross-engine decision rubric for self-hosting or recommending an LLM serving stack.

    457 GitHub stars~6.1k tokensUpdated 1 mo ago
    AI & LLM EngineeringAuto-check passed
  • Official

    Chooses the right serving container and current image URI for deploying a Hugging Face model to a SageMaker endpoint, preferring Hugging Face images over generic ones.

    11k GitHub starsUsed in 1 repo~4.6k tokens
    AI & LLM EngineeringAuto-check passed
  • Aider Delegate

    amElnagdy/delegate-skills

    Delegate a coding task to Aider (aider) as a background implementer, then review its diff and land it yourself.

    2.3k GitHub starsUsed in 2 repos~3k tokens
    AI & LLM EngineeringAuto-check passed
  • Dstack Prototyping

    dstackai/dstack

    Use with the dstack skill for model-serving work when the image, serving command, resources, backend/fleet choice, or service behavior is not proven.

    2.3k GitHub stars~1.6k tokensUpdated today
    AI & LLM EngineeringAuto-check passed

More from NVIDIA/OpenShell

All 23 skills in this repo
  • Official

    Maintain and validate OpenShell's build-only Windows MSVC lane for x64 and ARM64.

    15k GitHub stars~4.9k tokensUpdated today
    Auto-check passed
  • Create GitHub Issue

    NVIDIA/OpenShell

    Official

    Create GitHub issues using the gh CLI. An agent skill from NVIDIA/OpenShell.

    15k GitHub stars~1.7k tokensUpdated today
    Auto-check passed
  • Create GitHub PR

    NVIDIA/OpenShell

    Official

    Create GitHub pull requests using the gh CLI. An agent skill from NVIDIA/OpenShell.

    15k GitHub stars~1.9k tokensUpdated today
    Auto-check passed
  • Debug Openshell Cluster

    NVIDIA/OpenShell

    Official

    Debug why an OpenShell gateway deployment is unhealthy, unreachable, or unable to create sandboxes.

    15k GitHub stars~19k tokensUpdated today
    Auto-check: notes
  • Gator Gate

    NVIDIA/OpenShell

    Official

    Validate and monitor OpenShell GitHub issues and PRs using the gator: state machine.

    15k GitHub stars~19k tokensUpdated today
    Auto-check passed
  • Helm Dev Environment

    NVIDIA/OpenShell

    Official

    Start up, tear down, and configure the local Kubernetes development environment for OpenShell.

    15k GitHub stars~4.9k tokensUpdated today
    Auto-check passed

Questions about Debug Inference

What does Debug Inference do?

Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM. Debug Inference is an agent skill from NVIDIA/OpenShell, published by the product's own GitHub organization. Debug inference clients that use an attached provider and its native endpoint, including hosted APIs and host-local Ollama, vLLM, SGLang, TRT-LLM, LM Studio, or NIM.

When should I use Debug Inference?

Debug Inference fits situations like: provider attachment; endpoint policy; credential substitution; migration from the removed managed inference endpoint.

How do I install Debug Inference in Claude Code?

Run `npx skills add NVIDIA/OpenShell --skill debug-inference -a claude-code`. Or copy the skill folder (skills/debug-inference in NVIDIA/OpenShell) into .claude/skills/debug-inference in your project. Claude Code loads it when a task matches its description.

How do I install Debug Inference in Codex?

Run `npx skills add NVIDIA/OpenShell --skill debug-inference -a codex`. Or copy the skill folder (skills/debug-inference in NVIDIA/OpenShell) into .agents/skills/debug-inference in your project. Codex loads it when a task matches its description.

Can I use Debug Inference in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add NVIDIA/OpenShell --skill debug-inference -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/debug-inference, .gemini/skills/debug-inference, .github/skills/debug-inference and .opencode/skills/debug-inference in your project.

What does Debug Inference need to run?

SKILL.md names no scripts, command-line tools or credentials: Debug Inference is instructions for the agent only.

Does Debug Inference access the network?

SKILL.md names 1 domain. As links in the text: docs.nvidia.com. This is read from the text; nothing was executed.

Is Debug Inference safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Debug Inference use?

Debug Inference is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Debug Inference use?

About 1.9k tokens (SKILL.md is roughly 7.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Debug Inference?

Skills that share tags, products or a category with Debug Inference: Dev Bump (Netis/heron, 102 stars), Source Command Dev Docs Sync (Netis/heron, 102 stars), Agentsop LLM Engine Selection (agentsope/SkillAlchemy, 457 stars) and SageMaker Serving Image Selection (huggingface/skills, 11k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Debug Inference?

NVIDIA (a GitHub organization, an official publisher) maintains it in NVIDIA/OpenShell, which has 15,188 GitHub stars. The repository holds 23 skills in this directory. The repository was last updated on October 7, 2026.

Source: NVIDIA/OpenShell on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.