Skill Guide

How the daily tracker works.

⌄

$ cat ~/system/origin.log

> initializing...

What You're Looking At

The avatar you saw? It's not a static image. It's a Pillow-rendered badge generated from real activity data. Every day this system scans git commits, the content pipeline, and AI sessions. Scores them, grades them, and levels up an RPG character.

Here's how it all works.

$ ls ~/system/stack/

The Stack

Scanner

daily_scan.py · Python · stdlib · git

Auto-detects output from git commits, file modification times, and content directories. Classifies each file, computes weighted point scores, and merges into a structured JSON log. Safe to run multiple times per day.

Dashboard

daily_dashboard.py · Python · Pillow

Reads the daily JSON log and renders a visual receipt card. Header row with letter grade, stat boxes, platform breakdown bar, three-column accomplishments/pipeline/tokens layout, and a footer with summary stats. All monospace, all code-generated.

Progression Engine

progression_engine.py · Python · XP formulas

Takes daily output scores and converts them into RPG progression. XP accumulation, level-ups, title promotions, class specialization, and milestone unlocks. The bridge between "I shipped today" and "my character evolved."

Avatar Generator

avatar_generator.py · Python · Pillow · 6-tier system

Renders the RPG avatar badge you see on the landing page. Six visual tiers that evolve as you level up, from a blinking cursor placeholder to a fully rendered character with class insignia, XP bar, and title.

Skill Tree

_gen_skill_tree.py · Python · Pillow · sector visualization

Maps the entire GTM OS ecosystem into an RPG-style skill tree. Content engine, GTM operations, voice system, agent skills, infrastructure. Each node represents a real capability in the codebase.

$ git log --oneline ~/system/evolution

The Evolution

This didn't start as an RPG. It started as a question: "Can I prove I actually ship?"

Started as a Cursor skill prompt. "Scan my work and tell me what I shipped today." One command, one text summary.

Became a Python scanner + Pillow dashboard. The text summary wasn't enough. I wanted a visual receipt. So I wrote a scanner that outputs JSON and a renderer that draws a PNG. The daily log was born.

Realized the images were generated code, not static assets. Every pixel is computed from real data. The dashboard doesn't screenshot anything. It reads a JSON file and draws every element with coordinates, colors, and layout logic.

Added the RPG progression layer on top. If you're already computing daily output scores, why not accumulate XP? If you're already grading performance, why not assign titles? The progression engine turned a tracker into a game.

Now renders a character that evolves based on real output. The avatar you saw on the landing page? It's not a JPEG I uploaded. It's a Pillow-rendered badge that changes as the character levels up. Six tiers, each with distinct visual treatment.

The point: this is achievable. It's Python + Pillow + a few hundred lines of logic. No SaaS, no dashboard product, no API subscriptions. Just a codebase that tracks itself.

$ cat ~/system/titles.json

The Title Table

Every level range unlocks a new title and avatar tier. Here's the full progression from Terminal Initiate to Grand Master Cursor Slayer.

Level	Title	XP Required	Tier
1	Terminal Initiate	—	T1
5	Prompt Apprentice	300	T1
10	Repo Architect	1,200	T2
15	Pipeline Runner	3,000	T2
20	Context Weaver	6,000	T3
25	Skill Forger	11,000	T3
30	Voice Alchemist	18,000	T4
35	System Sovereign	30,000	T4
40	OS Architect	50,000	T5
45	Cursor Slayer	75,000	T5
50	Grand Master Cursor Slayer	120,000	T6

Class System

Your class is assigned based on your cumulative output distribution across all tracked days. The progression engine recalculates class on each daily scan. To shift class, you need sustained output in a new domain — a single day won't flip it. Each class applies a 1.5x XP multiplier to its primary domain while all other domains earn at the base 1.0x rate.

BuilderCode & Infrastructure

Focused on shipping. Code, scripts, infrastructure. 1.5x XP multiplier on all technical output.

Majority of cumulative points from monorepo builds, landing pages, feature systems, and code infra.

ScribeContent

Content-first. 1.5x XP multiplier on finalized posts, newsletters, and long-form writing.

Majority of cumulative points from finalized posts, newsletters, long-form writing, and content drafts.

StrategistGTM Operations

GTM operator. 1.5x XP multiplier on workflows, campaign copy, and partner operations.

Majority of cumulative points from workflows, campaign copy, partner operations, and outreach.

AlchemistEfficiency

AI-native. 1.5x XP multiplier on efficiency-rated output. Maximizes output per token dollar.

Highest cumulative efficiency rating (output per token dollar spent) across all tracked days.

PolymathBalanced

Balanced across all domains. 1.5x XP multiplier on the lowest-scoring domain to encourage balance.

No single domain exceeds 40% of cumulative output. Consistent spread across code, content, and GTM.

$ ./build --ready

Ready to build this?

Start with the basics. The daily tracker prompt gives you the scanner + dashboard. Everything you need to generate your own build receipts.

grab the prompt →
want the RPG layer too? →

← back to the log build your own →