Rea
Reverse engineer anything with agents, from app behavior down to native binaries.
A Claude agent that audits LLM apps and AI agents for token cost and rate-limit risk. It analyses usage logs, scans code for wasteful API calls, and returns a ranked savings plan. It can also apply the fixes, under a hard budget cap. Uses Claude Code subagents and a reusable skill, and every number comes from a script.
A Claude agent that audits an LLM application or AI agent for token cost and rate-limit risk. It returns a prioritised savings plan and, when you allow it, applies the fixes. AI Cost Optimizer audits applications built on large language models for avoidable token spend and rate-limit failures. It reads your usage logs and source code, finds the places where calls waste money (missing prompt caching, no output
git clone --depth 1 https://github.com/dutchy313/ai-cost-optimizer.git ~/.claude/skills/ai-cost-optimizer
Restart Claude Code afterwards. See the README for skill-specific setup.
Using Codex CLI? Copy the same folder to ~/.codex/skills/ai-cost-optimizer instead.
No specific triggers listed.
A Claude agent that audits an LLM application or AI agent for token cost and rate-limit risk. It returns a prioritised savings plan and, when you allow it, applies the fixes. AI Cost Optimizer audits applications built on large language models for avoidable token spend and rate-limit failures. It reads your usage logs and source code, finds the places where calls waste money (missing prompt caching, no output
Clone it into your personal skills folder: git clone --depth 1 https://github.com/dutchy313/ai-cost-optimizer.git ~/.claude/skills/ai-cost-optimizer — then restart Claude Code.
Ai Cost Optimizer is open source on GitHub (0 stars at last check) and published by dutchy313.
Browse other security skills in the Security category of this directory.