/ai--fine-tuning

/ai--fine-tuning is a Claude Code skill in the AI & Agents section. Decides honestly whether an LLM needs fine-tuning or better prompts, and if so prepares data, trains and compares before and after on held-out data.

Author's description

Fine-tune an LLM - decide vs prompting, prep dataset, train, eval before/after

How to ask for it

/ai--fine-tuning should I fine-tune or just improve my prompt?
/ai--fine-tuning prep my conversation dataset for training
/ai--fine-tuning compare model performance before and after fine-tuning

Install

curl -fsSL https://raw.githubusercontent.com/sgomez-dev/claude-skills/main/install.sh | bash

After installing with the script, type /ai--fine-tuning. Using Cursor, Windsurf or Codex? Platform guides

Demo coming soon

/ai--guardrailsLLM safety guardrails
Adds input and output filters, prompt-injection defense and personal-data handling around an AI feature, then proves them with adversarial tests.
Built here
/ai--llm-cost-optimizerCutting LLM costs
Measures where your app spends tokens, then applies model routing, caching and prompt compression, keeping each change only if an eval shows quality holds.
Built here
/ai--llm-evalLLM output evaluation
Sets up an evaluation suite for an AI feature: test cases, rubrics, code-based checks, LLM-as-judge with bias controls and a CI gate.
Built here
/ai--llm-observabilityLLM call monitoring
Instruments your AI app with traces, token and cost tracking, quality signals, dashboards and alerts for every model call.
Built here
/ai--mcp-serverBuilding an MCP server
Builds a Model Context Protocol server that exposes your project's functions, APIs or data to clients like Claude Code, with transport, auth and tests.
Built here
/ai--multi-agentMulti-agent system design
Designs a system of several AI agents with an orchestration pattern, contracts between agents and failure handling, or advises a single agent instead.
Built here

How to ask for it

Install

More from this section