llm-fine-tuning-lora

Production-grade skill package for Llm Fine Tuning Lora in data-mlops.

SKILL.md

GitHub

Llm Fine Tuning Lora (llm-fine-tuning-lora)

Domain: data-mlops
Progressive Disclosure: Loaded just-in-time when trigger conditions match.
Harness: Antigravity Native

Trigger Conditions

  • Activated when the task involves Llm Fine Tuning Lora or domain data-mlops.
  • Explicitly requested via /llm-fine-tuning-lora slash command or agent delegation.

Step-by-Step Instructions

  1. Analyze Context: Review existing code, configuration, and constraints before applying modifications.
  2. Formulate Plan: Outline the precise changes required, respecting project rules and token budget.
  3. Execute Incrementally: Apply atomic changes adhering to clean architecture and target language standards.
  4. Run Verification: Execute automated tests or compliance linters to validate correctness.
  5. Record Learning: Persist any newly discovered patterns or architectural decisions into memory.

Reference Resources & Checklist

  • Confirm inputs are sanitized and free of injection vulnerabilities.
  • Keep token footprint minimal; avoid unnecessary large file reads.
  • Ensure backward compatibility with existing project interfaces.

Verification Criteria

  • Automated tests pass with 0 failures.
  • Code conforms to relevant rules/ standards.
  • AgentShield security scan reports 0 Critical and 0 High severity issues.