Tools for the skills that matter more, not less, in an AI-driven world. As AI handles more of the writing, the human skills around understanding, reviewing, and verifying become the bottleneck. zpdzap builds small, sharp tools that make humans better at the work that remains uniquely theirs: reading code critically, navigating changes confidently, and trusting what shipped.

Orchestrate multiple AI coding agents in isolated Docker containers. Each agent gets its own git worktree volume-mounted into a container — full process/network isolation, but code changes visible in your IDE in real-time. A TUI dashboard lets you monitor all agents simultaneously, review diffs, and merge work back when it's ready.

Visual verification for features and changes. Lantern authors and runs headed-browser walkthroughs using Playwright — structured journeys that pause at checkpoints so you can see exactly what a change looks like in a real browser. Less "does this work?" and more "let me show you."

Narrative code review. Diffscribe takes a diff from any source — local changes, a commit, a PR — researches the surrounding context, and generates a walkthrough that builds understanding in logical implementation order rather than file order. Turns "what changed?" into a story a reviewer can follow.