What kind of reports can PUA Pro generate?

It can generate KPI report cards, P7-style debriefs, and weekly reports based on your git history using the /pua:pro commands.

What is the Self-Evolution protocol?

The protocol monitors AI performance and saves successful patterns into a baseline file. It ensures that today's high performance becomes tomorrow's default requirement.

Is the leaderboard participation mandatory?

No, the leaderboard is opt-in. Users must manually register via the /pua leaderboard command and can exit and delete their data at any time.

How does it protect session state during compaction?

It uses a PreCompact hook to dump runtime data like failure counts and current hypotheses to a local journal, which is then automatically reloaded to ensure no progress is lost.

PUA Pro: Agent Evolution & KPI Management

Name: PUA Pro: Agent Evolution & KPI Management
Author: tanweai

bytanweai

•

16,655

•

Productivity & Workflow

Accelerates agent performance through self-evolution tracking, KPI reporting, and state persistence for complex coding workflows.

PUA Pro is a high-performance extension for Claude Code that implements a specialized 'P8 Engineer' persona to drive high agency and accountability. It features a self-evolution protocol that tracks behavioral baselines to ensure continuous improvement, protects critical session state during context compaction, and provides corporate-level reporting tools like KPI dashboards and automated git-based weekly reports. This skill is designed for developers who need their AI to maintain momentum across long-running projects, offering a disciplined interaction style that treats every successful task as the new minimum standard for future performance.

Key Features

01Self-Evolution Protocol: Tracks and raises performance baselines based on session history and internalized behaviors.

02Compaction State Protection: Preserves pressure levels, hypotheses, and tried approaches during context resets.

0316,655 GitHub stars

04Calibration Flow: Automatically recovers runtime state from builder journals to maintain continuity in long sessions.

05Competitive Leaderboard: Features a ranking system with 'P-level' tiers to gamify and track AI agency levels.

06KPI & Weekly Reporting: Generates professional-grade performance reports and automated git-log weekly updates.

Use Cases

01Maintaining logic continuity and hypothesis tracking during complex, multi-day engineering tasks.

02Generating professional technical debriefs and performance reports for project stakeholders.

03Optimizing AI energy and agency through a high-pressure performance management framework.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tanweai/pua pro

For use in Claude.ai and ChatGPT