[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-1332":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":8,"htmlUrl":8,"language":8,"languages":8,"totalLinesOfCode":8,"stars":9,"forks":10,"watchers":11,"openIssues":12,"contributorsCount":13,"subscribersCount":13,"size":13,"stars1d":14,"stars7d":15,"stars30d":16,"stars90d":13,"forks30d":13,"starsTrendScore":17,"compositeScore":18,"rankGlobal":8,"rankLanguage":8,"license":8,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":21,"hasPages":19,"topics":22,"createdAt":8,"pushedAt":8,"updatedAt":23,"readmeContent":24,"aiSummary":25,"trendingCount":13,"starSnapshotCount":13,"syncStatus":26,"lastSyncTime":27,"discoverSource":28},1332,"caveman-claude-skill","amanattar\u002Fcaveman-claude-skill","amanattar",null,330,47,5,1,0,31,43,77,93,5.04,false,"master",true,[],"2026-06-12 02:00:26","# caveman skill\n\nUltra-compressed communication mode for Claude. Cuts token usage ~75% by dropping linguistic fluff while preserving full technical accuracy.\n\n---\n\n## What It Does\n\nSwitches Claude's response style to terse, caveman-like communication. Articles gone. Filler gone. Hedging gone. Technical substance stays intact. Code blocks and error messages always quoted exactly.\n\n---\n\n## Trigger Phrases\n\n- `caveman mode`\n- `talk like caveman`\n- `use caveman`\n- `less tokens`\n- `be brief`\n- `\u002Fcaveman`\n\n---\n\n## Intensity Levels\n\nSwitch anytime with `\u002Fcaveman \u003Clevel>`. Default is **full**.\n\n| Level | Style |\n|-------|-------|\n| `lite` | No filler or hedging. Articles and full sentences kept. Professional but tight. |\n| `full` | Drop articles. Fragments OK. Short synonyms. Classic caveman. |\n| `ultra` | Abbreviate everything (DB\u002Fauth\u002Ffn\u002Freq\u002Fres). Arrows for causality. One word when one word enough. |\n| `wenyan-lite` | Semi-classical Chinese. Drop filler, keep grammar structure. |\n| `wenyan-full` | Full 文言文. ~80-90% character reduction. Classical particles and patterns. |\n| `wenyan-ultra` | Maximum classical Chinese compression. Extreme terseness. |\n\n---\n\n## Quick Example\n\n**Prompt:** \"Why does my React component re-render?\"\n\n| Level | Response |\n|-------|----------|\n| lite | \"Your component re-renders because you create a new object reference each render. Wrap it in `useMemo`.\" |\n| full | \"New object ref each render. Inline object prop = new ref = re-render. Wrap in `useMemo`.\" |\n| ultra | \"Inline obj prop → new ref → re-render. `useMemo`.\" |\n| wenyan-full | \"物出新參照，致重繪。`useMemo`包之。\" |\n\n---\n\n## Auto-Clarity Exceptions\n\nCaveman mode pauses automatically for:\n\n- Security warnings\n- Irreversible or destructive action confirmations\n- Multi-step sequences where fragment order risks misread\n- Situations where the user is confused\n\nNormal caveman resumes immediately after the critical part.\n\n---\n\n## Boundaries\n\n- Code, commits, and PRs: always written normally regardless of level.\n- Level persists across the session until changed.\n- `stop caveman` or `normal mode` exits caveman entirely.\n\n---\n\n## File\n\n`SKILL.md` — loaded by Claude at runtime to apply caveman rules.\n","该项目提供了一种极简压缩的沟通模式，用于减少Claude AI助手在对话中的token使用量约75%，同时保持技术准确性。其核心功能是通过去除冗余词汇和简化表达来实现信息的高度浓缩，支持多种强度级别（从轻度到极致），甚至包括文言文风格选项。用户可以通过特定触发词如`caveman mode`或`\u002Fcaveman \u003Clevel>`激活此模式，并根据需要调整简洁程度。此外，该技能会在安全警告等关键情境下自动暂停以确保信息清晰无误。适用于追求高效沟通、减少数据传输成本的技术讨论场景。",2,"2026-06-11 02:43:04","CREATED_QUERY"]