Skills from clawhub.ai/paulpete

Use when testing Ralph's hat collection presets, validating preset configurations, or auditing the preset library for bugs and UX issues.

Unified TDD skill with three input modes — from spec, from task, or from description. Enforces test-first development using repository patterns, with proptest guidance and backpressure integration.

Lists all code tasks in the repository with their status, dates, and metadata. Useful for getting an overview of pending work or finding specific tasks.

Generates new Ralph hat collection presets through guided conversation. Asks clarifying questions, validates against schema constraints, and outputs production-ready YAML files.

Use when managing Ralph orchestration loops, analyzing diagnostic data, debugging hat selection, investigating backpressure, or performing post-mortem analysis

Dangerous skill: executes arbitrary code via `playwriter -e`, giving scripts access to the authenticated Chrome session and the Playwright Page API. It performs network navigation/interception (e.g., `https://example.com`, `**/api/**`) and can write files such as `/tmp/a11y.png`.

Transforms a rough idea into a detailed design document with implementation plan. Follows Prompt-Driven Development — iterative requirements clarification, research, design, and planning.

Use when bumping ralph-orchestrator version for a new release, after fixes are committed and ready to publish

Interactive terminal control via tmux for TUI apps, prompts, and long-running CLI workflows.

This skill documents creating terminal demo GIFs from asciinema recordings and converting/embedding them using `asciinema`, `agg`, and `svg-term`. It includes explicit shell commands (`asciinema rec`, `agg`), a network install (`cargo install --git https://github.com/asciinema/agg`), and a grep check for secrets in `demo.cast`.

This skill generates structured code task files and folder trees (e.g., `specs/{task_name}/tasks/`, `step01/task-01-*.code-task.md`) from descriptions or PDD plans. It accepts `file` paths and `URL` inputs and therefore must fetch remote content; it includes no instructions to read secrets or execute shell commands.