$ man context-wiki/python-for-gtm

Python is the glue language for automation. It connects things. It processes files. It generates images. It calculates stats. It runs on every operating system. And most importantly, Claude is excellent at writing it. You do not need to be a Python developer. You need to know enough to read scripts, understand what they do at a high level, and tell Claude what to build or fix. That is a different skill than writing Python from scratch. It is the difference between being a chef and being someone who can read a recipe and tell the chef what to change.

PATTERN

daily_scan.py scans git commits and content folders to calculate what I shipped today. It counts blog posts, drafts, published pieces, and partner deliverables. The output feeds into the daily tracker. daily_dashboard.py takes the scan results and generates a Pillow image card with stats, grades, pipeline info, and visual formatting. It creates the dashboard card that gets posted to track progress. rpg_sprites.py generates pixel art avatars for the RPG progression system. It creates idle animations, static sprites, and tool-specific character variants. 2,900 lines of Python that Claude wrote entirely from my descriptions. batch_rename.py processes files in bulk. Renaming, moving, reformatting. Anything that touches more than 10 files at once gets a script instead of manual work.

PRO TIP

Every Python script I use was built by Claude. The pattern is always the same. I describe what I want in plain English. Claude writes the code. I run it. If it works, done. If it breaks, I paste the error back to Claude and say "this broke, here is the error." Claude fixes it. I run it again. The script gets better with each iteration. This is the skill pattern applied to code. You do not need to debug Python yourself. You need to recognize when output is wrong and communicate what went wrong. Claude handles the actual programming. You handle the intent and the verification.

PATTERN

File processing: reading CSVs, parsing markdown, scanning directories. Every data pipeline starts with reading files. Image generation with Pillow: creating dashboard cards, branded images, social media assets. Pillow turns data into visuals without Photoshop. API calls with requests: hitting external APIs for data enrichment, posting to webhooks, pulling campaign stats. JSON manipulation: reading config files, transforming data structures, generating structured output for other tools. These four patterns cover 90% of the Python I use. If you understand what each does at a high level, you can direct Claude to build anything in the GTM automation space.