Skip to content
Skills Library
AI & agents · Tool integration

setup-guide

/setup-guide

Use when the user wants to set up inference-time scaling for the first time, or when its_hub is not yet installed/configured in the current environment.

Repository
its_hub ↗
Licence
Apache-2.0
Last repository change
28.09.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
3
Deprecation

No signals detected. Repository changed on 28.09.2026, without references to retired models or APIs.

Licence

Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.

What it does and when to use it

Use when the user wants to set up inference-time scaling for the first time, or when its_hub is not yet installed/configured in the current environment.

Original author description

Use when the user wants to set up inference-time scaling for the first time, or when its_hub is not yet installed/configured in the current environment.

How it works

  1. Step 1: Detect Environment
  2. Step 2: Install if Needed
  3. Step 3: Quick Setup or Custom
  4. Step 4: Collect Configuration
  5. Step 5: Ensure API Key
  6. Step 6: Save Config
  7. Adding More Models
Original SKILL.md headings · 115 lines, 645 words
SKILL.md preview
---
name: setup-guide
description: "Use when the user wants to set up inference-time scaling for the first time, or when its_hub is not yet installed/configured in the current environment."
allowed-tools: ["Bash(${CLAUDE_PLUGIN_ROOT}/scripts/its_detect.sh:*)"]
---

# its_hub Setup Guide

You are helping the user set up inference-time scaling.

## Step 1: Detect Environment

```!
"${CLAUDE_PLUGIN_ROOT}/scripts/its_detect.sh"
```

Also check for API keys in the environment:
```!
echo "openai_key=${OPENAI_API_KEY:+found}" "anthropic_key=${ANTHROPIC_API_KEY:+found}" "endpoint=${OPENAI_BASE_URL:-none}"
```

## Step 2: Install if Needed

If `library=missing`:
- Explain: "its_hub is a library for inference-time scaling — it generates multiple LLM responses and selects the best one using voting or scoring algorithms."
- Ask permission: "I can install it for you. Want me to proceed?"
- If yes and `installer=uv`: run `uv pip install "its_hub[lm]"`
- If yes and `installer=pip`: run `pip install "its_hub[lm]"`
- If `installer=none`: tell the user they need Python 3.11+ and pip/uv installed first

## Step 3: Quick Setup or Custom

**If an API key was detected**, offer a one-question fast path:

> "I detected your OpenAI API key. I can set up with these defaults:
> - Model: `gpt-4o-mini`
> - Algorithm: `self-consistency`
> - Budget: `8`
>
> Accept these defaults, or would you like to customize?"

If the user accepts, skip to Step 5 using:
- `endpoint`: `https://api.openai.com/v1` (or `OPENAI_BASE_URL` if set)
- `model`: `gpt-4o-mini`
- `algorithm`: `self-consistency`
- `budget`: `8`

**If no API key was detected**, or the user wants to customize, proceed to Step 4.

## Step 4: Collect Configuration

Ask these questions **one at a time**:

1. **Endpoint**: "What's your model endpoint URL?" — e.g., `http://localhost:8000/v1` for vLLM, `https://api.openai.com/v1` for OpenAI
2. **Model name**: "What's the model identifier?" — e.g., `gpt-4o`, `Qwen/Qwen2.5-32B-Instruct`
3. **Algorithm**: "Which scaling algorithm do you want to use?" — consult the `inference-scaling-guide` skill for guidance if the user is unsure.
   - **Self-consistency** — Votes on the most common answer. No extra setup needed.
   - **Best-of-N** — Scores each with an LLM judge. Requires a judge model.
4. **Budget**: "Which budget do you want to use?" — consult the `inference-scaling-guide` skill for guidance if the user is unsure.

### Algorithm-Specific Config

**Self-consistency:**
- Ask: "Do you need regex patterns for answer extraction? (e.g., `\boxed{...}` for math). If unsure, skip — default voting works for most cases."
- Ask: "Will your prompts involve tool/function calls?" If yes, ask which voting strategy: `tool_name`, `tool_args`, or `tool_hierarchical`

**Best-of-N:**
- Ask: "Which model should be the judge? (This can be the same model or a different one)"
- Collect: judge model name, judge endpoint (default: same as generation model)

## Step 5: Ensure API Key

API keys are read from environment variables — **never store them in the config file**.

If no API key was detected in Step 1, tell the user to set the appropriate environment variable:

> "Set your API key as an environment variable before running scaling:
> ```bash
> export OPENAI_API_KEY="sk-..."        # OpenAI or OpenAI-compatible endpoints
> export ANTHROPIC_API_KEY="sk-ant-..."  # Anthropic endpoints
> ```
> For local vLLM endpoints that don't require authentication, no API key is needed."

## Step 6: Save Config

Write the config to `.its-hub/config.json`:

```json
{
  "models": {
    "default": {
      "endpoint": "<endpoint>",
      "model": "<model_name>"
    }
  },
  "algorithm": "<algorithm>",
  "budget": "<budget>",
  "algorithm_config": {}
}
```

Add `.its-hub/` to `.gitignore` if not already present.

Confirm the config file was written, then report success:

> "Setup complete! To run scaling, use the `inference-scaling` skill.
>
> **API keys** are read from environment variables, not the config file. Make sure the appropriate variable is set in your shell:
> ```bash
> export OPENAI_API_KEY="sk-..."        # OpenAI or OpenAI-compatible endpoints (including vLLM with auth)
> export ANTHROPIC_API_KEY="sk-ant-..."  # Anthropic endpoints
> ```
> Local endpoints (e.g., vLLM without auth) don't need an API key."

## Adding More Models

If this skill is invoked again and a config already exists, ask: "You already have a configuration. Do you want to update it or add another model?"

If adding a model: collect endpoint and model name. Add a new entry to the `models` dict using the model name as key.

Versions

RepositoryScoreLicence
This page · red-hat-ai-innovation-team/its_hub100Apache-2.0
SchulzWill/weekpost100MIT
red-hat-ai-innovation-team/training_hub100Apache-2.0
red-hat-ai-innovation-team/sdg_hub100Apache-2.0
Azure/sap-automation-qa100MIT
DojoCodingLabs/remotion-superpowers90MIT

Database snapshot: 13.11.2026 · package version ff2559ea3dde (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100inference-scaling
    Use when the user wants to run inference-time scaling on a prompt — detect environment, execute scaling, and present results. For algorithm selection, budget tuning, reward models, and troubleshooting, consult the inference-scaling-guide skill.
  • 100inference-scaling-guide
    Guides users through inference-time scaling with its_hub, including algorithm selection (Self-Consistency, Best-of-N, Beam Search, Particle Filtering), budget tuning, reward model setup, tool-calling integration, interpreting results, and troubleshooting. Use when the user is…
  • 100setup
    Provisions the oracle ML inference daemon with onnxruntime via uv. Use when setting up local ONNX model inference for skill quality evaluation.
  • 100streaming-inference-setup
    Configure streaming inference setup operations. Auto-activating skill for ML Deployment. Triggers on: streaming inference setup, streaming inference setup Part of the ML Deployment skill category. Use when working with streaming inference setup functionality. Trigger with…
  • 100configuring-auto-scaling-policies
    Configure use when you need to work with auto-scaling. This skill provides auto-scaling configuration with comprehensive guidance and automation. Trigger with phrases like "configure auto-scaling", "set up elastic scaling", or "implement scaling".
  • 100configuring-auto-scaling-policies
    Configure use when you need to work with auto-scaling. This skill provides auto-scaling configuration with comprehensive guidance and automation. Trigger with phrases like "configure auto-scaling", "set up elastic scaling", or "implement scaling".

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.