VibeRPGSimulator

Pin the bug, watch it live

Hunter · Testing and debugging3 spells in orderProposed, not testedUnrated

Play the combo

Foe: Flaky Wisp, a failing or flaky test.

Foes

  • Flaky Wisp

Party · Context

  • HunterYour agent100/100

  1. Flaky Wisp

    The Flaky Wisp blocks the way.

    What really happensIt stands for a failing or flaky test.

    Hunter: Context 100/100.

  2. Flicker

    The Flaky Wisp uses Flicker.

    What really happensA flaky test passes on one run and fails on the next.

    Hunter takes 8. Hunter: Context 92/100.

  3. Matt Pocock TDD

    A scroll that makes the hero write the failing test before the fix.

    What really happensWrites the regression test first, through public interfaces, one behavior at a time.

    Flaky Wisp takes 33.

  4. Flicker

    The Flaky Wisp uses Flicker.

    Hunter takes 8. Hunter: Context 84/100.

  5. Anthropic Webapp Testing

    A scroll that sends a headless browser through a local app.

    What really happensExercises the UI with Python Playwright scripts against your local app.

    Flaky Wisp takes 33.

  6. Flicker

    The Flaky Wisp uses Flicker.

    Hunter takes 8. Hunter: Context 76/100.

  7. Chrome DevTools MCP

    An orb that opens live Chrome so the hero can inspect the page.

    What really happensInspects console, network and performance evidence in Chrome, after the other browser is done.

    Flaky Wisp takes 34. Flaky Wisp falls.

  8. Victory

    The Flaky Wisp falls.

    What really happensWrite a test that pins the behavior, exercise the UI, then read the runtime evidence. Use the browser tools one at a time instead of running competing browser instances.

Cast order

  1. Matt Pocock TDD

    Writes the regression test first, through public interfaces, one behavior at a time.

  2. Anthropic Webapp Testing

    Exercises the UI with Python Playwright scripts against your local app.

  3. Chrome DevTools MCP

    Inspects console, network and performance evidence in Chrome, after the other browser is done.

Why it works

Write a test that pins the behavior, exercise the UI, then read the runtime evidence. Use the browser tools one at a time instead of running competing browser instances.

Records

Seen together0sessions containing every step
In this order0sessions with matching first-use order
Test recoveries0 of 0sessions with a failing test

Curator's machine: macOS 27.0.1 · Apple Silicon (arm64), read when these stats were published. Harness versions installed then: Codex 0.154.0, Claude Code 2.1.282, OMP 18.8.3. Session logs do not record the OS.

Co-occurrence and order describe the logs; they do not show that this combo caused an outcome.

Learn the spells

  1. 1. Matt Pocock TDD

    Claude Code

    claude plugin install mattpocock-skills@claude-plugins-official

    Codex

    codex plugin marketplace add mattpocock/skills
codex plugin add mattpocock-skills@mattpocock

  2. 2. Anthropic Webapp Testing

    Claude Code

    /plugin marketplace add anthropics/skills
/plugin install example-skills@anthropic-agent-skills

  3. 3. Chrome DevTools MCP

    Claude Code

    claude mcp add chrome-devtools --scope user npx chrome-devtools-mcp@latest

    Codex

    codex mcp add chrome-devtools -- npx chrome-devtools-mcp@latest