pawel-cell

Use when building software with AI agents and you need a serious end-to-end workflow instead of vibe coding. Covers harness choice, context discipline, cleanup, review loops, launch pressure, and security basics.

Use after an AI-built feature works but the code has duplicated mechanics, repeated API calls, or messy structure. Guides a cleanup pass that extracts reusable service-layer modules without changing behavior.

Use when you have a small PR or feature and want an agent to repeatedly fix review feedback until tests pass and the PR is merge-ready. Works with AI reviewers, Greptile-style review, or human feedback.

Use when an AI coding agent is guessing API names or framework behavior from incomplete docs. Shows how to provide local source-code references so the agent searches the real implementation before coding.

Generate today's daily note in an Obsidian vault by carrying forward yesterday's open todos, summarizing active projects, and adding an "on deck" section. Use for morning briefings, scheduled daily-note creation, or when the user asks for "today's note".

Scan an Obsidian vault's Inbox/ folder, classify each note, file it into the right destination folder, add tags, and link to existing project pages. Use for periodic vault cleanup or when the user dumps rough notes and wants them organized.

Generate images via OpenRouter using direct chat-completions calls (gpt-5.4-image-2, gpt-5-image, gemini-3-pro-image, etc.) — POST to /chat/completions with modalities=["image","text"], decode the base64 data URL, save to disk, return the file path.

Drive the cmux native macOS terminal app from CLI or socket — workspaces, panes, surfaces, browser automation, notifications, sidebar metadata, session restore. Use whenever the user mentions cmux, wants to control terminal layout from an agent, automate browser panels on macOS, send notifications/flashes to the sidebar, or integrate an AI agent with cmux hooks. macOS only (14.0+).

Use when someone wants to learn software development by reading, modifying, and debugging real working code instead of studying abstract theory. Turns any repo or code snippet into a practical step-by-step exercise.