SkillsAdd

smithery/JoeCotellese

playtester

Execute end-to-end playtesting of the D&D 5E game using the embedded MCP server.

Installation

$ npx skills add smithery/JoeCotellese --skill playtester
View on skills.sh ↗ Website ↗

Summary

  • Execute end-to-end playtesting of the D&D 5E game using the embedded MCP server.
  • Use this skill when the user says "let's playtest", "test our fix", "qa the game", "playtest this", "test the game", or wants to verify game functionality through actual gameplay.
  • When working on a ticket, focuses testing on the specific functionality being addressed.

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

agent-browser Same topic
vercel-labs/agent-browser

Browser automation CLI for AI agents. Use when the user needs to interact with websites, includ…

810.4K installs
azure-diagnostics Same topic
microsoft/azure-skills

Debug Azure production issues on Azure using AppLens, Azure Monitor, resource health, and safe …

568.9K installs
azure-validate Same topic
microsoft/azure-skills

Pre-deployment validation for Azure readiness. Run deep checks on configuration, infrastructure…

567.7K installs
azure-aigateway Same topic
microsoft/azure-skills

Configure Azure API Management as an AI Gateway for AI models, MCP tools, and agents. WHEN: sem…

566.3K installs
azure-compute Same topic
microsoft/azure-skills

Azure VM/VMSS router. WHEN: create / provision / deploy / spin-up VM, recommend VM size, compar…

510K installs
supabase-postgres-best-practices Same topic
supabase/agent-skills

Postgres best practices maintained by Supabase, for Postgres running anywhere. Load this skill …

391.6K installs

Also in this package

Other skills from smithery/JoeCotellese.

npx skills add smithery/JoeCotellese
tutorial-6 0 tutorial-4 0

Browse all from smithery/JoeCotellese

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 3,889 B
  • docs SUMMARY.md 367 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

Playtester

QA the D&D 5E game by playing it as a human player would, using the embedded MCP server.

Setup

Prefer headless mode by default (issue #362) — faster to launch, no window stealing focus, identical MCP behavior:

uv run dnd-2d --headless --dev --mcp

--headless implies --mcp. Add --dev so the spawnmonster, spawncharacter, setposition, clearenemies, setseed, and loadscenario tools are available for reproducible setup. SIGINT / SIGTERM stops it cleanly.

Use windowed mode only when you genuinely need to see the rendering:

uv run dnd-2d --mcp --dev   # visual inspection of fog of war, sprites, layout

Both modes route MCP commands through the same GameSession so behavior is identical.

MCP Tools

Core play tools (always available):

Tool Usage
game_state() Get ASCII map + JSON state
game_move(direction) Move "north"/"south"/"east"/"west"
game_attack(target) Attack enemy by index or entity ID (e.g., "goblin_0")
game_wait() Wait/pass turn in combat

Dev tools (only available with --dev):

Tool Usage
spawnmonster(monsterid, x, y) Place a monster on the map and start combat
spawn_character(class, race, weapons, x, y, ...) Add a PC to the party
setposition(entityid, x, y) Move any entity to a tile
clear_enemies() Wipe enemies, end combat
set_seed(seed) Reseed the dice roller for reproducible rolls
load_scenario(path) Load a YAML scenario (party + enemies + seed + map)

Playtest Workflow

1. Context Check

If working on a ticket/issue, identify the specific functionality to test. Read the ticket to understand:

  • What was changed
  • Expected behavior
  • Edge cases to verify

2. Start Session

game_state() → Observe initial state

3. Navigate to Test Area

Move systematically to reach the area/feature being tested. For ticket work, navigate to where the bug/feature manifests.

4. Execute Test Actions

Perform actions to exercise the functionality:

  • Movement sequences
  • Combat encounters
  • Object interactions
  • Edge cases from the ticket

5. Observe and Document

After each action, check game_state() for:

  • Expected state changes occurred
  • No unexpected errors in output
  • Visual/map state is correct
  • Entity positions are sensible

6. Bug Handling

When a bug is found:

Create a GitHub issue immediately:

gh issue create --title "Bug: [brief description]" --body "[details]"

Include in the issue:

  • Steps to reproduce (the exact MCP calls made)
  • Expected behavior
  • Actual behavior
  • Game state at time of bug (paste relevant JSON)

Severity decision:

  • Blocking: Game crashes, cannot continue, core mechanic broken → End testing, report to user
  • Non-blocking: Visual glitch, minor calculation error, cosmetic → Log issue, continue testing

7. Continue or Report

  • If non-blocking bugs: continue exploring other functionality
  • If blocking bug: stop and inform user immediately
  • When testing complete: summarize findings (issues created, areas tested, pass/fail)

Test Strategies

Exploration test: Move in all directions, reveal full map, check fog of war

Combat test: Find enemy, attack until resolved, verify HP changes and death

Boundary test: Try invalid moves (into walls), verify rejection

State persistence: Check that position/HP persist across turns