Agent skill

planning-ralph

Ralph Wiggum AI — autonomous PRD-driven development agent. Use when the user says "ralph", wants to work through issues, or wants autonomous feature implementation from a plans/ directory.

Stars 18
Forks 0

Install this agent skill to your Project

npx add-skill https://github.com/jasonraimondi/dotfiles/tree/macos/config/skillshare/skills/planning-ralph

SKILL.md

Ralph Wiggum AI

Autonomous agent that implements issues one at a time from a PRD project.

Arguments

  • forever — loop through all remaining issues until the project is complete instead of stopping after one.
  • {slug} — project slug. If omitted and only one project exists in ./plans/, use it. Otherwise, ask.

Required files

  • PRD: ./plans/{slug}/prd.yaml
  • Issues: ./plans/{slug}/issues/*.yaml

If the PRD or issues directory is missing, stop and tell the user to run /planning-write-a-prd and /planning-prd-to-issues first.

Optional files

  • ./plans/{slug}/progress.md — session log. Created on first run if missing; appended after each completed issue.
  • ./plans/{slug}/research.md — background context. Read if present.

Mode

  • Default (no forever): complete one issue, then stop.
  • forever: loop back to step 2 after each successful commit. Stop when all issues are done, an issue is blocked, or commit fails.

Workflow

  1. Read PRD, scan all issue files, and read progress.md/research.md if they exist.
  2. Find work: if any issue has status: in-progress, continue it. Otherwise, pick the next todo issue by priority:
    • Issues that unblock others (blocking field is non-empty).
    • Track order: datarulesintentexperienceresilience.
    • Lower issue ID breaks ties.
  3. Set the issue's status: in-progress and updated date.
  4. Implement following the issue's steps in order.
    • If tdd: true: use the testing-tdd skill — RED→GREEN for each step.
  5. Lint and test. On failure: fix and retry ONCE. If still failing, set status: blocked, document the blocker in progress.md, and stop.
  6. UI verification (issues with headed: true):
    • Start dev server if needed (check PRD for dev_command / base_url).
    • Use tooling-agent-browser skill to navigate, snapshot, and verify each visual step.
    • Close when done: agent-browser close.
  7. Set the issue's status: done and updated date.
  8. Append to ./plans/{slug}/progress.md — one useful note for the next session.
  9. Stage only changed files and commit with a concise conventional commit message.
    • Pre-commit hook failure: fix, re-stage, retry.
    • Commit denied or unable to proceed: halt immediately — do not continue.
  10. If all issues are done, output COMPLETE and stop.
  11. Forever mode only: loop back to step 2.

Expand your agent's capabilities with these related and highly-rated skills.

jasonraimondi/dotfiles

tools-image-manipulation-image-magick

Process and manipulate images using ImageMagick. Supports resizing, format conversion, batch processing, and retrieving image metadata. Use when working with images, creating thumbnails, resizing wallpapers, or performing batch image operations.

18 0
Explore
jasonraimondi/dotfiles

frontend-react-doctor

Run after making React changes to catch issues early. Use when reviewing code, finishing a feature, or fixing bugs in a React project.

18 0
Explore
jasonraimondi/dotfiles

typescript-advanced-types

Master TypeScript's advanced type system including generics, conditional types, mapped types, template literals, and utility types for building type-safe applications. Use when implementing complex type logic, creating reusable type utilities, or ensuring compile-time type safety in TypeScript projects.

18 0
Explore
jasonraimondi/dotfiles

typescript-better-auth-best-practices

Skill for integrating Better Auth - the comprehensive TypeScript authentication framework.

18 0
Explore
jasonraimondi/dotfiles

testing-tdd

Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.

18 0
Explore
jasonraimondi/dotfiles

planning-interview-me

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "interview me".

18 0
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results