- Home
- Skills
- Game Development
- Unity Playtest: Automated Game Testing
Works with the AI Agents you already use
Unity Playtest: Automated Game Testing
Claude plays your Unity build for you: real key presses, host + client with latency, and a bug report at the end.
Free
Unity Playtest: Automated Game Testing
Example session with this skill installed
Check if the player gets stuck in any colliders in the current scene and verify if the 'Interact' key works on the 'Chest' prefab in Play Mode.
- Read your context and instructions
- Compiled the unity playtest
I have initialized the playtest environment and executed the scene probes followed by a solo play session.
Playtest Report: Scene Audit & Interaction Test
| Case | Layer | Result | Evidence |
|---|---|---|---|
| Collider Probe | Probe | FAIL | 2 objects found overlapping static geometry: PlayerSpawn_02, Props/Crate_04. |
| Interact Logic | Solo | PASS | PtKey(Interact) called; Chest.IsOpen transitioned to True within 0.1s. |
| Console Check | Probe | PASS | No new NullReferenceExceptions or errors recorded during interaction. |
| Visual State | Human | PENDING | Script confirmed state change; person must verify if the lid animation looks smooth. |
Observations
- The
PlayerSpawn_02point is partially inside a wall collider, which will cause the character controller to glitch on spawn. - The
Crate_04mesh is intersecting with the floor plane by 0.15 units. - Interaction with the
Chestprefab is functional at the code level.
Next steps
- Move
PlayerSpawn_02at least 0.5 units away fromEnvironment/Wall_North. - Adjust the Y-position of
Props/Crate_04to0.0. - Open the
MainLevelscene to visually confirm the chest animation plays correctly.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual playtesting in Unity is time-consuming and prone to human error, especially when verifying network sync or physics edge cases. This skill automates the QA process by driving the Unity Editor through the Pipeline CLI, allowing you to run scripted input, probe scenes for broken references, and validate multiplayer sessions with simulated latency.
What it does
- Scene probing identifies missing references, objects stuck in colliders, and invalid spawn points without saving scene changes.
- Scripted input simulates real key presses and actions through the Input System to verify gameplay loops and feature logic.
- Multiplayer testing orchestrates host and client sessions using Netcode for GameObjects to detect races and sync issues.
- Security audits attempts to forge client intents to ensure the server correctly refuses unauthorized or impossible requests.
- Automated reporting generates a structured pass/fail report with evidence, separating scripted proofs from human feel.
How it works
- Initialization runs the
doctor.pyscript to verify the Unity CLI, the Pipeline package, and initialize the git-ignored.playtest/environment. - Recon & Mapping identifies the project's input actions and session logic to populate a project-specific adapter.
- Execution runs probes for static analysis or enters Play Mode to execute scripted steps against the Editor and standalone builds.
- Cleanup closes sessions, deletes temporary builds, and reverts project settings to ensure zero side effects in git.
Frameworks & tools
Requires Unity with the com.unity.pipeline package installed. Scripts are compatible with Python 3.8+ and utilize the Unity Pipeline CLI for Editor communication.
Why this beats prompting it yourself
Standard LLMs cannot "see" a running Unity scene or interact with the Input System. This skill provides the specific C# probes and Python drivers needed to actually execute code inside the Editor, verify physics states, and manage multiplayer lifecycle events that a text-only prompt would miss.
Use cases
- Verifying a PR's "How to test" section by running the exact steps in Play Mode.
- Finding elusive physics bugs where players get stuck in geometry or spawn points.
- Testing late-joiner synchronization and network variable replication under high latency.
- Validating that server-side authority correctly blocks cheated client requests.
Known limitations
Cannot judge "feel," sound quality, or aesthetic fun. Requires the Unity Pipeline package to be manually installed in the project first. Builds with the playtest runtime must not be shared.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Game Development.
Cannot judge game feel or aesthetics. Requires manual installation of com.unity.pipeline. Scripts require Python 3.8+. Multiplayer testing requires creating temporary development builds.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 19 days ago
- Free with an Agensi account
- Software
Git