zig-sdl3-bindings

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation references an external repository for its core dependency at git+https://codeberg.org/7Games/zig-sdl3. This reference is standard for the skill's purpose.
  • [COMMAND_EXECUTION]: The skill provides an API (sdl3.Process) to spawn and interact with external system processes, as demonstrated in the references/system-platform.md documentation.
  • [DATA_EXFILTRATION]: The skill documents capabilities to read and write files across the system (sdl3.filesystem), access sensitive folders such as the user home directory, interact with the system clipboard (sdl3.clipboard), and perform network communications (sdl3.net).
  • [PROMPT_INJECTION]: The skill describes an attack surface for indirect prompt injection due to its ability to ingest untrusted data from various sources and perform privileged system operations.
  • Ingestion points: Data can enter the agent context via the clipboard (references/clipboard.md), network sockets (references/net.md), and file reading operations (references/filesystem-io.md).
  • Boundary markers: The provided code patterns do not include specific delimiters or instructions to ignore embedded commands in processed data.
  • Capability inventory: The skill facilitates subprocess execution, file system modification, and network egress.
  • Sanitization: The reference patterns do not demonstrate explicit sanitization of input data before it is used in logic or system calls.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 09:31 PM