A new Claude skill applies the Earned vs Burned Framework to measure AI software delivery by real business outcomes rather than effort metrics like tokens or lines of code. The framework scores tasks on a five-level hierarchy, from backlog to verified production impact, and calculates metrics such as Earn Rate and Earned per AI Token—a metric the body notes the industry lacks. The skill integrates with major project management tools and is designed for teams managing traditional, outsourced, or AI-agent-hybrid workflows.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Harveer Singh released an open-source Claude skill that scores software delivery tasks on a five-level outcome hierarchy — from backlog (0) to verified KPI impact in production (1.0) — and calculates metrics like Earn Rate, Earned/Hours, and Earned per AI Token to replace traditional velocity reporting.
Why it matters
Teams building with AI typically measure effort (token usage, lines of code, story points) rather than actual business value. This framework redefines success as tangible outcomes — revenue impact, user confirmation, KPI movement — making it possible to tell whether AI work actually earned its cost, which is particularly relevant for companies managing outsourced or hybrid human-AI delivery.
What to watch
The skill integrates with Linear, Asana, GitHub Issues, Jira, and Azure DevOps; it can be installed directly into Claude Desktop or Cowork via a .skill file, and works for FTE teams, outsourced delivery, and AI-agent workflows. The framework is Singh's IP, but the skill code is MIT licensed.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
