Score 12 agent autonomy dimensions on a 1–5 rubric and map totals to ZTE levels.
npx skills add https://github.com/CleanExpo/Pi-Dev-Ops --skill leverage-audit
Please help me install this Agent Skill. Skill: leverage-audit Source: https://github.com/CleanExpo/Pi-Dev-Ops/tree/main/skills/leverage-audit Command: npx skills add https://github.com/CleanExpo/Pi-Dev-Ops --skill leverage-audit
This Skill helps teams diagnose where agent autonomy gains will deliver the highest ROI by turning qualitative design concerns into a consistent, scored “12 Leverage Points” assessment.
Ask an agent to score your current autonomy pipeline against the 12 leverage dimensions, then recommend the three highest-ROI fixes based on the lowest scores.
High-intent search queries and answers about installing and using this skill.
Agent autonomy evaluation uses a 12 leverage points rubric to score dimensions like context precision, tool availability, and error recovery on a 1–5 scale. Pinpointing the three lowest scores reveals which workflow fixes yield the highest ROI.
A ZTE level is a maturity label—Manual, Assisted, Autonomous, or Zero Touch—derived from your total autonomy score. It communicates your current agent workflow capability and standardization progress across the 12 evaluated dimensions.
Scoring agent workflows involves applying a rubric-driven 1–5 scale across 12 dimensions including feedback loops, session continuity, and cost efficiency. Comparing these scores highlights spec quality and trigger automation gaps causing workflow bottlenecks.
Tooling readiness is a core dimension in autonomy evaluation, scoring tool availability on a 1–5 scale. Assessing integration alongside model selection and knowledge retention determines if your pipeline can support autonomous triggers.
Prioritizing context strategy fixes involves scoring context precision and spec quality against 10 other dimensions. Treating the bottom three lowest-scoring dimensions as high-ROI targets ensures prompt and context improvements directly increase workflow maturity.
floflo777Evaluate RAG systems using retrieval, generation, and latency metrics.
zero-rehqClassify task descriptions into 12 domains with confidence scores and routing recommendations.
jayhjenkinsAutomate three-tier QA testing with TEST_INDEX.md and RUNS/ reports.
az9713Evaluate repository harness maturity with a 12-control rubric and generate a Markdown report.
