The Agent Skills Directory

[SAFE]: The skill acts as a technical resource for Unity developers to manage in-app purchasing. It provides detailed guides for migrating from legacy systems and integrating new monetization features using official Unity APIs.
[INDIRECT_PROMPT_INJECTION]: The skill contains logic for scanning a project's source code, manifest files, and configuration assets (such as IAPProductCatalog.json, .ucat, and .csv files) to assess the environment and provide tailored instructions. This constitutes an attack surface where malicious data in those files could theoretically attempt to influence the agent's output. However, given the technical nature of the task and the lack of autonomous execution of these instructions, the risk is minimal and consistent with the skill's utility.
Ingestion points: The skill instructions prompt the agent to read content from Assets/, Packages/, and ProjectSettings/ to identify third-party packages and native billing code.
Boundary markers: The skill does not define specific delimiters for untrusted file content.
Capability inventory: The skill provides templates for C# code and configuration file modifications, but does not autonomously execute dangerous system commands.
Sanitization: The skill focuses on generating valid code templates for the user to review and implement.
[EXTERNAL_DOWNLOADS]: The documentation references official Unity packages (e.g., com.unity.purchasing) and well-known services such as Stripe and Coda. These references are documented neutrally and align with the skill's primary purpose.
[COMMAND_EXECUTION]: The skill describes standard development tasks, including the use of Unity's batchmode and package management via the Unity Editor, which are routine operations for the target use case.

implement-in-app-purchases