AEL Dashboard for OpenClaw Users

Turn by turn real-time session monitoring that tells you exactly when your AI agent should spawn a fresh session — and how much money you're leaving on the table if it doesn't. Built on the Agent Efficiency Limit framework. Patent pending.

• Per-turn cost tracking with cache read/write breakdown
• Minimum savings/turn — live cost of irrelevant context
• Spawn recommendation with 2 user set thresholds
• Cache TTL timer (Anthropic 5min / OpenAI 60min)
• Turn cost and minimum savings/turn charts
• Works with Anthropic, OpenAI, and Grok
• Single HTML file — no server, no install, no account
• Free updates

Works with:

Available on:

Full AEL monitoring for your existing agent framework. Each package includes the framework adapter, sidecar engine, and visual dashboard — bundled and ready to run. Add a few lines to your agent code and get real-time cost tracking, context drag monitoring, and spawn recommendations out of the box.

• Per-turn cost tracking with cache read/write breakdown
• Live context drag and spawn recommendations
• Visual dashboard — single HTML file, no server needed
• Works on Windows and Mac
• Free updates

Choose your framework:

Available on:

AEL for non-developers. Takes simple user inputs to calculate the health status of your current AI session, delivers reset recommendations based on the AEL framework, and generates a continuity handoff so your next session picks up right where you left off. Works in any browser — no API key required.

• Works directly in your browser — no setup needed
• Monitors Claude.ai, ChatGPT, and Gemini sessions
• Estimates context size and tracks turns
• Nudges you when it's time to reset
• Free updates

Works with:

Available on:

A drop-in wrapper for direct OpenAI and Anthropic API calls. Tracks token usage across turns, computes the cost of carrying irrelevant context, and tells you when starting a fresh session is cheaper than continuing. No dashboard required — built for developers who want AEL in their own code.

• Drop-in replacement for OpenAI and Anthropic API clients
• Tracks V_shed and S_real per turn with no setup
• Spawn recommendations built into every response
• Custom context component tracking and auto-shed
• Tunable S_real to match your agent's actual setup size
• Free updates

Works with:

Available on: