Skip to main content

Prompt Optimization

Skill Verified Active

Applies prompt repetition to improve accuracy for non-reasoning LLMs

Purpose

To improve the accuracy of non-reasoning LLMs on structured tasks by automatically repeating their prompts.

Features

  • Applies prompt repetition to increase LLM accuracy
  • Automatic activation for Haiku agents on specific tasks
  • Configurable repetition count and enablement
  • Zero latency impact due to prefill stage operation
  • Provides performance metrics and configuration options

Use Cases

  • Improving accuracy for unit tests
  • Enhancing reliability of linting and formatting tasks
  • Increasing precision in parsing and data extraction
  • Ensuring accuracy in list operations (find, filter, count)

Non-Goals

  • Improving accuracy for reasoning LLMs (Opus, Sonnet)
  • Tasks outside of structured operations
  • Introducing latency penalties

Installation

npx skills add asklokesh/loki-mode

Runs the Vercel skills CLI (skills.sh) via npx — needs Node.js locally and at least one installed skills-compatible agent (Claude Code, Cursor, Codex, …). Assumes the repo follows the agentskills.io format.

Quality Score

Verified
100 /100
Analyzed about 19 hours ago

Trust Signals

Last commit3 days ago
Stars922
LicenseNOASSERTION
Status
View Source

Similar Extensions

© 2025 SkillRepo · Find the right skill, skip the noise.