Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Registers a CacheForge account, connects an upstream LLM provider, and returns a tenant API key. All OpenAI-compatible tools route through the CacheForge proxy after a single config change.
Auto-detects the upstream provider from the API key prefix and provisions the full proxy config in one command, replacing manual account setup and config editing.
When to use it
- Routing Claude Code requests through CacheForge to reduce token spend
- Connecting OpenRouter or Fireworks AI as an upstream behind a single endpoint
- Onboarding a new workspace to a shared CacheForge tenant
- Validating proxy connectivity before running production agent workloads
- Switching upstream providers without changing tool configurations
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Set up CacheForge with my OpenAI key
- 1Checks environment for OPENAI_API_KEY and detects 'sk-' prefix as an OpenAI-compatible upstream
- 2Collects user email and password for CacheForge registration
- 3Runs python3 setup.py provision to register the account and mint a tenant API key
- 4Runs python3 setup.py validate to confirm a test request routes successfully through the proxy
- 5Prints the OpenClaw provider snippet and applies it to ~/.openclaw/openclaw.json with user approval
CacheForge tenant API key (cf_...) issued and OpenClaw configured to route requests through https://app.anvil-ai.io
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.