Find the ghost tokens. Fix them. Survive compaction. Avoid context quality decay.

codex cost-optimization cost-management tokenomics token-usage context-window cost-tracking claude-code token-optimization token-efficiency context-engineering claude-plugin session-continuity token-optimizer pi-coding-agent pi-extension hermes-plugin tokenoptimization smart-compaction claude-code-mod

• Updated Nov 12, 2026
• Python

ContextIQ is an open-source framework for LLM token optimization, context engineering, deterministic prompt compression and AI agent cost reduction across GitHub Copilot, Cursor, Claude Code, Windsurf, and VS Code.

vscode cursor cost-optimization windsurf githubcopilot llm prompt-engineering token-usage cost-tracking claude-code token-optimization context-engineering tokenoptimization

• Updated Aug 20, 2026
• Python

Human steers. Agent builds. Stop burning AI agent credits on wasted context.

cli productivity opensource ai devtools developer-tools cursor copilot claude prompt-engineering contextengineering tokenoptimization codewithai

• Updated Jun 9, 2026
• TypeScript

The open source coding agent, enhanced with 3 token saving features like multi model routing and dynamic auto compaction

ai localllm aiengineering codingagents tokenoptimization tokenop

• Updated Oct 30, 2026
• TypeScript