skills.rest
leogurja
leogurjaCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

caveman-setup

Wire repository LLM callsites through the Caveman gateway for spend measurement.

1|Updated May 3, 2021
Search Tags:#integration#llm#observability#openai#anthropic#gateway#spend tracking
One-click install
npx skills add https://github.com/leogurja/dotfiles --skill caveman-setup-leogurja
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman-setup
Source: https://github.com/leogurja/dotfiles/tree/main/home/dot_agents/skills/caveman-setup
Command: npx skills add https://github.com/leogurja/dotfiles --skill caveman-setup-leogurja

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams shipping LLM-powered code have no visibility into what each request costs. This Skill integrates a repository's existing LLM callsites with the Caveman gateway so every request is measured and priced, without changing application behavior. ## Core Features & Use Cases - Callsite Discovery: Scans dependency files and source code for LLM SDKs (OpenAI, Anthropic, Vercel AI SDK, LangChain, LiteLLM) and raw provider HTTP calls before making any change. - Gateway Wiring: Rewrites each callsite's base URL to the gateway path with an app slug and adds the required authentication headers, keeping secrets in environment variables only. - Verified Integration: Sends one real verification request and reports the actual HTTP status and token usage, with explicit failure templates for unreachable gateways, invalid keys, and routing errors. - Use Case: A developer with a Python app using the Anthropic SDK asks to "set up caveman"; the Skill finds the client, points its base URL at the gateway, stores the API key in the repo's env file, and confirms measurement with a live request. ## Quick Start Ask the agent to set up caveman in this repository using the gateway URL, API key, provider key mode, and dashboard URL from your setup prompt.

Frequently Asked Questions about caveman-setup

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add LLM spend tracking to my app?▼

Point each LLM client's base URL at the Caveman gateway with an app slug and add the x-cave-api-key header. The gateway measures every request in record mode without changing request or response bytes.

Which LLM SDKs can route through the Caveman gateway?▼

OpenAI, Anthropic, Vercel AI SDK, LangChain/LangGraph, LiteLLM, and raw HTTP calls to provider APIs are supported with published wiring recipes. Frameworks like google-genai, crewai, and pydantic-ai follow integration docs fetched from the docs origin.

Does the gateway change my LLM requests or responses?▼

No. Record mode is byte-preserving: it measures what the app sends and what it costs, and changes nothing else. No optimization is enabled and verified savings remain $0 until an optimizer is explicitly turned on.

Where should the Caveman API key be stored?▼

Store CAVE_API_KEY in the repo's existing env file such as .env or .env.local, never in source code. If that file is not gitignored, add it to .gitignore before writing the key.

Why does gateway verification return a 401 or 404 error?▼

A 401 cave_invalid_api_key means the gateway rejected the key; mint a new one from the dashboard and update the env file. A 404 cave_route_not_found usually means a malformed app slug or a path that does not match the SDK's protocol.

Related Skills

Explore More Software Engineering Skills→
mnfstmnfst
7.4k

manifest

Route OpenClaw LLM requests to cost-effective models by query complexity.

Official
Advanced
PostHogPostHog
75

exploring-llm-costs

Aggregate LLM spend and token metrics from PostHog events.

Official
Advanced
cosmicstack-labscosmicstack-labs
364

monitoring-observability

Establish an observability stack correlating metrics, logs, and distributed traces.

Official
Intermediate
curiositechcuriositech
2

reverse-proxy-for-agents

Route and secure AI agent HTTP, WebSocket, and LLM-router traffic with reverse proxy configurations.

Official
Advanced