context-compactorToken-based context compaction for local models (MLX, llama.cpp, Ollama) that don't report context limits.
Install via ClawdBot CLI:
clawdbot install emberdesire/context-compactorGrade Good — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Potentially destructive shell commands in tool definitions
rm -rf ~Calls external URL not in known-safe list
https://github.com/E-x-O-Entertainment-Studios-Inc/openclaw-context-compactorAI Analysis
The skill's primary function is local context management with no evidence of data exfiltration or credential harvesting. The only external reference is a GitHub repository URL for installation, which aligns with the skill's stated purpose and poses minimal risk.
Audited Apr 17, 2026 · audit v1.0
Generated Mar 11, 2026
Law firms use local models to analyze lengthy legal documents and case histories. Context Compactor prevents silent truncation, ensuring accurate summaries of older case details while maintaining recent legal arguments in full context for coherent analysis.
Researchers employ local LLMs for literature reviews and data synthesis. The plugin proactively summarizes older research papers, allowing models to focus on recent findings and queries without hitting token limits, improving research efficiency.
Companies deploy local models for handling customer support tickets with long conversation histories. Context Compactor summarizes earlier interactions, enabling the AI to retain key customer details while responding to new issues within token constraints.
Development teams use local models to review extensive codebases and commit histories. The plugin compacts older code discussions, ensuring the AI can analyze recent changes and provide relevant feedback without exceeding context limits.
Healthcare providers utilize local LLMs to process patient histories and medical notes. Context Compactor summarizes older records, allowing models to focus on recent symptoms and treatments for accurate diagnostic support while adhering to privacy with local deployment.
Offer the Context Compactor as a free open-source plugin to drive adoption. Generate revenue through paid support plans, including customization, tuning for specific models, and priority troubleshooting for enterprise clients using local AI deployments.
Partner with local model providers (e.g., MLX, llama.cpp) to bundle Context Compactor as a premium feature. Charge licensing fees per integration or offer it as part of higher-tier packages that enhance model reliability and user experience.
Develop a cloud-based SaaS platform that extends Context Compactor with advanced analytics, multi-model support, and automated tuning. Target AI development teams with subscription tiers based on usage volume and features like real-time monitoring and optimization insights.
💬 Integration Tip
Start by installing via the npx setup command and adjust maxTokens based on your model's context size; use /context-stats to monitor usage and fine-tune settings for optimal performance.
Scored May 17, 2026
Use CodexBar CLI local cost usage to summarize per-model usage for Codex or Claude, including the current (most recent) model or a full model breakdown. Trigger when asked for model-level usage/cost data from codexbar, or when you need a scriptable per-model summary from codexbar cost JSON.
Check Antigravity account quotas for Claude and Gemini models. Shows remaining quota and reset times with ban detection.
使用豆包（火山引擎）语音合成大模型 API 将文本转换为语音音频文件。支持声音复刻音色（S_ 开头的音色ID）和官方预置音色。当用户要求"语音合成"、"文字转语音"、"TTS"、"朗读文本"、"生成语音"、"用我的声音读"、"豆包语音"、"声音复刻合成"等相关请求时，务必使用此 skill。即使用户只是说"帮我把...
Intelligent model routing for sub-agent task delegation. Choose the optimal model based on task complexity, cost, and capability requirements. Reduces costs...
自动生成科技新闻摘要。从多个来源（RSS、Twitter、GitHub、Web Search）抓取科技新闻，整合后生成摘要。
Sync OpenRouter models used by OpenClaw into this installation's config. Fetches the OpenClaw app leaderboard from OpenRouter, verifies model IDs against the...