The Agent Skills Directory

[SAFE]: The skill implements a camera presentation grammar tool for game development. Its core functionality involves a Python script that validates authored camera paths and properties defined in JSON files.
[DATA_EXPOSURE_AND_EXFILTRATION]: The provided validation script, boom_sweep.py, implements security best practices including a 4 MiB file size limit for input processing and a refusal to overwrite existing files when generating reports. The script does not utilize any network-capable modules or access sensitive system directories.
[DYNAMIC_EXECUTION]: The skill avoids unsafe data handling by using a strict JSON parser that rejects duplicate keys and non-finite constants. No dynamic code evaluation or runtime code generation from user-supplied data is performed.
[COMMAND_EXECUTION]: The skill uses standard CLI invocation for its validation tool. The Python script is self-contained, using only standard libraries, and does not execute shell commands or spawn sub-processes.
[INDIRECT_PROMPT_INJECTION]: Although the skill processes external configuration data (JSON), the script performs rigorous schema validation, type checking, and range validation, which prevents the data from influencing the agent's execution flow in an unsafe manner.

camera-shots