Soak Test Protocol
Generates a checklist for long play sessions to catch memory leaks, crashes and fatigue
- Repository
- Donchitos/Claude-Code-Game-Studios
- Stars
- 32,965
- Licence
- MIT
- Files
- 1 file
What it does
When invoked, the agent builds a structured protocol for an extended game play session (30 minutes to several hours) with timed checkpoints and items to observe, such as memory growth, frame rate drift, crashes, audio and input problems, and player fatigue. It reads project settings and design documents to tailor the checklist, then writes a markdown report file for a human tester to fill in. The human still does the actual playing; the agent prepares the protocol and analysis structure.
Great for
- Game teams in a Polish phase preparing for a release gate check
- Developers verifying that a memory or stability fix holds up over time
- QA testers who need a consistent, timed checklist for long sessions
Skip it if
- Quick smoke checks or single-feature playtests, which the skill itself describes as shorter and different (the skill is designed for extended runs)
- Projects that are not game projects or lack an engine or performance budget, since much of the guidance depends on those settings (fallbacks are unclear)
Try asking
- “Run a 2h soak test focused on memory for my Godot project before the release gate”
- “Create a 4h balance soak protocol to check whether combat gets repetitive in my RPG”
- “Make a 30m stability soak checklist for the new inventory scene”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
soak-test.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip soak-test.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip soak-test.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Set the duration and focus arguments explicitly, for example 2h memory, so the protocol matches what you want to test.
- Record the memory unit your editor displays at T+0 and keep using it, since the skill compares growth against the session's own baseline.
- Fill in the checkpoint results live during the run; the protocol is only a template until a human plays and observes.
What's inside
- SKILL.md
Redistributed unmodified from Donchitos/Claude-Code-Game-Studios under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.