VibeRPGSimulator

Red-test bug hunt

Quest: Reproduce a bug, fix it and keep a test that guards it

Proposed, not tested

Equipment

Hunter Codex
Quest
Reproduce a bug, fix it and keep a test that guards it
Hero
CodexThe agent that runs the quest
Class
Hunter · Testing and debugging
Weapon
GPT-6 SolWarhammer: Built for coding and agent workMana rate 11 per 1k tokens
Charge
HighReasoning effort
Author
The curator

Sentry shows the crash, the debugging scroll finds the cause before any fix, TDD pins the bug with a failing test first, and Playwright reproduces it in a browser when the bug is in the UI.

Can't be appraised: Proving Grounds can't equip orbs (MCP servers).

Skill tree

  1. ScoutGather context

  2. CraftMake the thing

    Nothing equipped here

  3. TestCheck the work

  4. ShipDeliver it

    Nothing equipped here

Quest replay

Quest: "Reproduce a bug, fix it and keep a test that guards it". Four waves, then the Backlog Dragon.

Foes

  • Fog Wraith

Party · Context

  • CodexHunter100/100

  1. Wave 1 of 4 · Scout

    The Fog Wraith steps up.

    What really happensGather context

    Codex: Context 100/100.

  2. Haze

    The Fog Wraith breathes haze. Swings start to miss.

    What really happensWithout current docs, the agent guesses at APIs that may have changed.

    Codex takes 3. Codex: Context 97/100. Codex: Haze.

  3. Sentry MCP

    An orb that pulls live Sentry issues into the fight.

    What really happensScope the URL with an organization slug, or org and project, so tools default there and extra discovery tools stay hidden.

    Fog Wraith takes 50.

  4. Haze cleared

    Fresh pages burn the haze away.

    What really happensCurrent docs replace guesses with facts.

  5. Haze

    The Fog Wraith uses Haze.

    Codex takes 3. Codex: Context 94/100. Codex: Haze.

  6. Warhammer

    Codex swings the Warhammer.

    What really happensA guess at an outdated API fails.

    Fog Wraith avoids the hit.

  7. Haze

    The Fog Wraith uses Haze.

    Codex takes 3. Codex: Context 91/100.

  8. Warhammer

    Codex swings the Warhammer.

    Fog Wraith takes 10. Fog Wraith falls.

  9. Wave 2 of 4 · Craft

    The Error Imp steps up.

    What really happensMake the thing

  10. Crash

    The imp hurls a stack trace.

    What really happensThe running app throws an error the agent has to trace.

    Codex takes 6. Codex: Context 85/100.

  11. Warhammer

    Codex swings the Warhammer.

    Error Imp takes 50 (critical hit). Error Imp falls.

  12. Wave 3 of 4 · Test

    The Bug Slime and the Flaky Wisp step up.

    What really happensCheck the work

    Flaky Wisp: Flicker.

  13. Flicker

    The Flaky Wisp uses Flicker.

    Codex takes 3. Codex: Context 82/100.

  14. Superpowers Systematic Debugging

    A scroll that bars fixes until the hero has a root cause.

    What really happensPhase 1 requires the error text, a reliable repro, and a look at recent changes. No repro means more data, no guess.

    Flaky Wisp avoids the hit.

  15. Pinned

    Pinned in place, it cannot slip away.

    What really happensA test that runs the same way each time shows the real failure.

    Bug Slime: Pinned. Flaky Wisp: Pinned.

  16. Regress

    The Bug Slime uses Regress.

    Codex takes 3. Codex: Context 79/100.

  17. Flicker

    The Flaky Wisp uses Flicker.

    Codex takes 3. Codex: Context 76/100.

  18. Matt Pocock TDD

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

    What really happensIt proposes seams, the public interfaces where tests observe behavior, and waits for you to confirm them.

    Flaky Wisp takes 40. Flaky Wisp falls.

  19. Regress

    The Bug Slime uses Regress.

    Codex takes 3. Codex: Context 73/100.

  20. Microsoft Playwright MCP

    An orb that opens a live browser so the hero can click through your app.

    What really happensEach step returns an accessibility snapshot, a structured list of page elements, so the agent works without reading pixels.

    Bug Slime takes 45. Bug Slime falls.

  21. Wave 4 of 4 · Ship

    The Gate Warden steps up.

    What really happensDeliver it

    Gate Warden: Barrier.

  22. Boss · Backlog Dragon

    The Backlog Dragon guards the quest: "Reproduce a bug, fix it and keep a test that guards it".

    What really happensIt stands for the whole quest, every task on the list.

  23. Warhammer

    Codex swings the Warhammer.

    Gate Warden is behind a barrier.

  24. Scope creep

    The dragon breathes scope creep across the party.

    What really happensNew asks pile onto a task while it is underway.

    Codex takes 5. Codex: Context 68/100.

  25. Bar the gate

    The warden raises the gate. Nothing passes until the checks are green.

    What really happensA release waits until tests, builds and reviews pass.

  26. Warhammer

    Codex swings the Warhammer.

    Gate Warden is behind a barrier.

  27. Claw

    The Backlog Dragon uses Claw.

    Codex takes 6. Codex: Context 62/100.

  28. Bar the gate

    The Gate Warden swings the gate shut.

    Codex takes 5. Codex: Context 57/100.

  29. Warhammer

    Codex swings the Warhammer.

    Gate Warden takes 43.

  30. Gate open

    The checks turn green and the gate opens.

    What really happensTests, build and review pass, so the release can go.

  31. Scope creep

    The Backlog Dragon uses Scope creep.

    Codex takes 5. Codex: Context 52/100.

  32. Warhammer

    Codex swings the Warhammer.

    Gate Warden takes 17. Gate Warden falls.

  33. Claw

    The Backlog Dragon uses Claw.

    Codex takes 6. Codex: Context 46/100.

  34. Warhammer

    Codex swings the Warhammer.

    Backlog Dragon takes 45.

  35. Scope creep

    The Backlog Dragon uses Scope creep.

    Codex takes 5. Codex: Context 41/100.

  36. Warhammer

    Codex swings the Warhammer.

    Backlog Dragon takes 47.

  37. Call the backlog

    The dragon roars, and another task crawls out of the backlog.

    What really happensFinishing one task often surfaces the next one on the list.

  38. Warhammer

    Codex swings the Warhammer.

    Bug Slime takes 25. Bug Slime falls.

  39. Scope creep

    The Backlog Dragon uses Scope creep.

    Codex takes 5. Codex: Context 36/100.

  40. Warhammer

    Codex swings the Warhammer.

    Backlog Dragon takes 62 (critical hit).

  41. Enraged

    Wounded, the dragon rages.

    What really happensThe last stretch of a big job is often the hardest.

    Backlog Dragon: Enraged.

  42. Claw

    The Backlog Dragon uses Claw.

    Codex takes 8. Codex: Context 28/100.

  43. Warhammer

    Codex swings the Warhammer.

    Backlog Dragon takes 65 (critical hit).

  44. Scope creep

    The Backlog Dragon uses Scope creep.

    Codex takes 6. Codex: Context 22/100.

  45. Compact

    The hero rolls the scroll of memory tight and fights on.

    What really happensWhen context runs low, the agent summarizes the session so far and keeps going. Small details can get lost.

    Codex recovers 28. Codex: Context 50/100.

  46. Warhammer

    Codex swings the Warhammer.

    Backlog Dragon takes 1. Backlog Dragon falls.

  47. Quest clear

    The Backlog Dragon falls. Quest complete.

    What really happensEvery equipped spell did its part of "Reproduce a bug, fix it and keep a test that guards it".

After the battle

  • Codex took 17 turns to clear four waves.
  • Context ran low once, so the agent compacted it.
  • No Craft or Ship spell, so those waves ran on basic attacks.
  • Codex spawns subagents when you ask for them, or when AGENTS.md or a skill tells it to.

Copy this loadout

Prompt for Codex

Paste this into Codex. It installs each spell, leaves API keys and sign-ins to you, and ends with the command that restarts it with this weapon and charge.

Set up these tools for me in Codex.
Loadout: https://vibe-rpg-simulator.vercel.app/loadouts/red-test-bug-hunt/

1. Sentry MCP
   Source: https://github.com/getsentry/toolkit
   Run:
     codex mcp add sentry --url https://mcp.sentry.dev/mcp

2. Superpowers Systematic Debugging
   Source: https://github.com/obra/superpowers/blob/main/skills/systematic-debugging/SKILL.md
   I type in Codex:
     /plugins

3. Matt Pocock TDD
   Source: https://github.com/mattpocock/skills/blob/main/skills/engineering/tdd/SKILL.md
   Run:
     codex plugin marketplace add mattpocock/skills
     codex plugin add mattpocock-skills@mattpocock

4. Microsoft Playwright MCP
   Source: https://github.com/microsoft/playwright-mcp
   Run:
     codex mcp add playwright npx "@playwright/mcp@latest"

How to do it:
- Skip anything that is already installed. Run the rest in order and check that each step worked.
- Commands under "I type in Codex" are mine to enter. List them for me at the end, in order.
- Never ask me to paste an API key or token into this chat. When a command needs one, such as YOUR_API_KEY, give me the full command to run in my own terminal.
- Ask me for any other value you can't work out, such as a project path.
- If a command asks a question or needs a browser sign-in or an open app, tell me what to do and wait.
- If a command fails, follow the install steps in the tool's source for Codex instead. Show me any command that is not on this list before you run it, and install nothing else.

When you're done, tell me what is ready and what still needs me. New tools load in a new session, so finish by telling me to restart Codex with GPT-6 Sol and reasoning effort high:
codex --model gpt-6-sol --config model_reasoning_effort='"high"'

Terminal commands

Or paste these lines into a terminal. The first part installs every spell for Codex; the last line starts it with the same weapon and charge. Lines starting with # are notes or commands you type inside the agent.

# Loadout: Red-test bug hunt
# Quest: Reproduce a bug, fix it and keep a test that guards it
# Hero: Codex · Weapon: GPT-6 Sol · Charge: High
# From https://vibe-rpg-simulator.vercel.app/loadouts/red-test-bug-hunt/

# 1. Learn the spells
# Scout: Sentry MCP
codex mcp add sentry --url https://mcp.sentry.dev/mcp
# Test: Superpowers Systematic Debugging
# in Codex: /plugins
# Test: Matt Pocock TDD
codex plugin marketplace add mattpocock/skills
codex plugin add mattpocock-skills@mattpocock
# Test: Microsoft Playwright MCP
codex mcp add playwright npx "@playwright/mcp@latest"

# 2. Start the hero with this weapon and charge
codex --model gpt-6-sol --config model_reasoning_effort='"high"'

Edit a copy