What problem does it solve?
This Skill addresses redundant token consumption during AI agent work sessions that drives up operational costs and reduces efficiency without delivering value, while ensuring quality and correctness are never sacrificed in the process.
Core Features & Use Cases
- FR-26 Rule Enforcement: Implements core execution-time economy rules including scoped reads, no re-reads of unchanged in-context files, two-strike retry rule, and targeted edits, each paired with quality guards to prevent correctness compromises.
- Token Waste Auditing: Provides a path to run /token-waste-audit on Claude Code or use repo-side fallback metrics on other surfaces to identify redundant consumption candidates mapped to FR-26 rules.
- Context Compression Discipline: Extends economy rules to large inputs and outputs with guidelines for routing, extracting, and preserving retrieval paths for large artifacts to avoid unnecessary context loading.
- Use Case: Use this Skill when debugging long tool-using AI development sessions, responding to user questions about high session costs or token waste, or running formal token waste audits for your AI workflow.
Quick Start
Invoke the token-economy skill when you need to audit your last AI development session for redundant token usage and receive a report of potential waste candidates aligned with FR-26 economy rules.
