Applies prompt repetition to improve accuracy for non-reasoning LLMs
安装
npx skills add asklokesh/loki-mode --skill prompt-optimizationnpx skills add asklokesh/loki-mode --skill prompt-optimization -a claude-code -gnpx skills add asklokesh/loki-mode --skill prompt-optimization -a codex -ggemini skills install https://github.com/asklokesh/loki-mode.git --path examples/agent-skills/prompt-optimizationgh skill install asklokesh/loki-mode prompt-optimizationgit clone --depth 1 https://github.com/asklokesh/loki-mode /tmp/asklokesh__loki-mode
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/asklokesh__loki-mode/examples/agent-skills/prompt-optimization ~/.claude/skills/prompt-optimization # Claude Code
cp -R /tmp/asklokesh__loki-mode/examples/agent-skills/prompt-optimization ~/.agents/skills/prompt-optimization # Codex, Cursor, Gemini CLI, Copilot, OpenCode