genai-hub-foundation-models — Claude Code skill for SAP models
The genai-hub-foundation-models Claude Code skill lists foundation models in SAP AI Core GenAI Hub and shows their details. It runs when a model name or executable ID is needed for a deployment.
Apache-2.0 256 installs ★ 153 updated 2026-10-22
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a claude-codeWhat genai-hub-foundation-models does
The skill fetches the current foundation-model catalog from SAP AI Core GenAI Hub. It can show providers, versions, capabilities, and context length, with filters for executable ID or scenario and a detailed JSON output.
It requires Python 3.11+, uv, and sap-ai-sdk-core. If credentials are not set up, call aicore-admin-resources first. When a model name is uncertain, the skill checks the live catalog; a name already supplied by the user skips that lookup.
The skill finds the exact model name for create_llm_deployment and provides curl examples for deployed models. It does not create or manage deployments; that work goes to aicore-lifecycle-management.
What it helps with
- browse available models by provider
- find a model's executable ID
- check versions, capabilities, or context length
- get the exact model name for
create_llm_deployment - retrieve a
curlexample for a deployed model
Example prompts for the agent
- List the available Google models and their executable IDs.
- Find a model that supports streaming and show its context length.
- Give me a
curlexample for calling an already deployed model.
How to install genai-hub-foundation-models
1Skills CLI (Vercel)
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a claude-code2Manually: put the files in place
~/.claude/skills/genai-hub-foundation-models/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/genai-hub-foundation-models/
1Skills CLI (Vercel)
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a codex2Manually: put the files in place
.agents/skills/genai-hub-foundation-models/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a cursor2Manually: put the files in place
.cursor/skills/genai-hub-foundation-models/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install SAP/ai-skills-library plugins/aicore-skills/skills/genai-hub-foundation-models2Manually: put the files in place
.github/skills/genai-hub-foundation-models/SKILL.md1Skills CLI (Vercel)
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a gemini-cli2Platform CLI
gemini skills install https://github.com/SAP/ai-skills-library --path plugins/aicore-skills/skills/genai-hub-foundation-modelsManually: put the files in place
.devin/skills/genai-hub-foundation-models/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about genai-hub-foundation-models
How do I install genai-hub-foundation-models in Claude Code?
Run npx skills add SAP/ai-skills-library --skill genai-hub-foundation-models -a claude-code. The skill file is placed at ~/.claude/skills/genai-hub-foundation-models/SKILL.md.
What does genai-hub-foundation-models do?
It lists foundation models from SAP AI Core GenAI Hub, including their versions, capabilities, context lengths, and executable IDs. It can also provide request examples for deployed models.
What does genai-hub-foundation-models require?
It requires Python 3.11+, uv, sap-ai-sdk-core, and configured credentials. If credentials are not set up, call aicore-admin-resources first.
Source text
SKILL.md · Apache-2.0
---
name: genai-hub-foundation-models
description: >
List and explore foundation models available in SAP AI Core GenAI Hub: browse supported
providers, model names, versions, capabilities, and executable IDs needed to create deployments.
WHEN: user wants to see what models are available, find an executable-id or model-name for
deployment, browse providers (Azure OpenAI, Anthropic, Google, Mistral, Cohere, etc.), check
model capabilities or context length, get the correct model-name to pass to create_llm_deployment,
or get curl inference examples for a deployed model (chat completions, Responses API, embeddings,
image generation, etc.).
DO NOT USE FOR: creating or managing deployments — use `aicore-lifecycle-management`.
compatibility: Requires Python 3.11+, uv, and sap-ai-sdk-core.
allowed-tools: Bash, Read
---
## Rules
1. Fetch the live model list when the model name is unknown or unconfirmed — skip if the user already named a specific model.
2. `executable-id` and `model-name` values from this skill are the inputs needed by `aicore-lifecycle-management` to create a deployment.
3. If credentials are not configured, invoke `aicore-admin-resources` first.
4. Before executing any curl example, get a Bearer token with:
```bash
export TOKEN=$(uv run skills/aicore-admin-resources/scripts/get_token.py)
```
Never add `2>&1` or pipe the command — that prints the raw JWT into the conversation.
---
## List Available Models
```bash
# Show all options and examples:
uv run scripts/list_foundation_models.py --help
# One provider only (e.g. Azure OpenAI):
uv run scripts/list_foundation_models.py --executable-id azure-openai
# Detailed view (versions, capabilities, context length, streaming support):
uv run scripts/list_foundation_models.py --response-format detailed
# Machine-readable JSON for scripting:
uv run scripts/list_foundation_models.py --json | jq '.[].model'
uv run scripts/list_foundation_models.py --response-format detailed --json | jq '.[] | select(.model == "gpt-4o") | .versions'
# Models supported by the Orchestration service only:
uv run scripts/list_foundation_models.py --scenario-id orchestration
```
---
## Provider → Executable ID Mapping
| Provider | `--executable-id` |
| -------------- | ------------------- |
| Azure OpenAI | `azure-openai` |
| Anthropic | `aws-bedrock` |
| Google | `gcp-vertexai` |
| Amazon Bedrock | `aws-bedrock` |
| Mistral AI | `aicore-mistralai` |
| Cohere | `aicore-cohere` |
| Perplexity | `perplexity-ai` |
| NVIDIA | `aicore-nvidia` |
| Open Source | `aicore-opensource` |
| SAP | `aicore-sap` |
Supported models in your instance may differ from this table — use the script to get the live list.
---
## Create a Deployment
Once you have the `executable-id` and `model-name`, check whether a RUNNING deployment for that model already exists:
```bash
uv run skills/aicore-lifecycle-management/scripts/get_deployments.py \
--scenario foundation-models --status RUNNING --json
```
If one is found, use `AskUserQuestion` to ask whether to reuse it or create a new one. Then invoke `aicore-lifecycle-management` with both values.
The `modelName` value must exactly match the catalog output — a mismatch causes an "Invalid Configuration" error at deployment time, not at configuration creation.
---
## When to Load Reference Files
| Trigger | Load |
| -------------------------------------------------------------------------------------------- | ---------------------- |
| Curl inference examples, provider-specific endpoint paths, or about to call a deployed model | `references/MODELS.md` |
---
## Finding the Deployment URL for a Curl Call
1. **Get the `executable-id`** — look it up in the Provider → Executable ID table above.
2. **Find the running deployment:**
```bash
uv run skills/aicore-lifecycle-management/scripts/get_deployments.py \
--scenario foundation-models --executable-id <executable-id> --status RUNNING --json
# use .deployment_url and .resource_group from the matching entry
```
3. See `references/MODELS.md` for export setup and curl examples per provider.
---
## Handoffs
- **Credential/auth error?** → invoke `aicore-admin-resources`
- **Ready to create a deployment with the model you found?** → invoke `aicore-lifecycle-management`
© SAP · Apache-2.0 · SAP/ai-skills-library