The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to configure build.zig.zon to fetch the raylib-zig dependency directly from a remote Git repository (https://github.com/raylib-zig/raylib-zig#main). This process involves downloading and building external code during the compilation phase, which is a standard part of the Zig ecosystem but introduces a dependency on a third-party source.
[INDIRECT_PROMPT_INJECTION]: The skill documents numerous functions that ingest untrusted external data, creating a surface for indirect prompt injection.
Ingestion points: Functions such as rl.loadTexture, rl.loadSound, rl.loadModel, rl.loadDroppedFiles, and UI widgets like rg.textBox or rl.getClipboardText ingest data from the filesystem, network-adjacent sources, or direct user input.
Boundary markers: The provided instructions do not include explicit boundary markers or warnings to the agent to treat data from these sources as untrusted.
Capability inventory: The skill enables several potentially dangerous actions including filesystem writes (rl.takeScreenshot, rl.exportMesh, rl.exportWave), and network navigation via rl.openURL.
Sanitization: No explicit sanitization or validation logic is provided for the content ingested through these interfaces before it is processed or used to influence agent behavior.
[COMMAND_EXECUTION]: The skill documentation relies on the Zig build system (zig build), which executes build scripts (build.zig) that can perform arbitrary operations on the host system as part of the project compilation and dependency resolution process.

zig-raylib