{"skill":{"slug":"subagent-context-compactor","displayName":"Context Compactor","summary":"上下文压缩代理，采用分层压缩策略，基于内存使用触发机制。处理HOT/WARM/COLD三层数据，优化token使用。当用户需要压缩对话上下文、优化内存使用、管理会话历史、减少token消耗时使用此技能。特别适用于长时间对话、复杂任务处理、需要保留重要历史信息的场景。","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":311,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1773106509002,"updatedAt":1777527536600},"latestVersion":{"version":"1.0.0","createdAt":1773106509002,"changelog":"Initial native agent release with hierarchical compression and memory-based triggers.\n\n- Migrated from plugin/JS implementation to standalone Python agent with API server and monitoring tools\n- Introduced HOT/WARM/COLD three-level context compression strategy\n- Automatic compaction triggered by token usage, message counts, timer, or manual command\n- New monitoring and reporting scripts for compression status and resource usage\n- Added configuration files for compression thresholds and strategies\n- Removed previous JS/TypeScript plugin files; all logic is now in Python and shell scripts","license":"MIT-0"},"metadata":null,"owner":{"handle":"geoshan","userId":"s174dnjfwwym02wderbkkxkt8n85b7an","displayName":"geoshan","image":"https://avatars.githubusercontent.com/u/10161131?v=4"},"moderation":{"isSuspicious":true,"isMalwareBlocked":false,"verdict":"suspicious","reasonCodes":["suspicious.exposed_secret_literal","suspicious.llm_suspicious"],"summary":"Detected: suspicious.exposed_secret_literal, suspicious.llm_suspicious","engineVersion":"v2.4.5","updatedAt":1777527536600}}