The Agent Skills Directory

[DATA_EXFILTRATION]: The skill directs the agent to read and source the sensitive .env file using the commands grep -q '^WORLDLABS_API_KEY=.' .env and set -a; . .env; set +a. This practice exposes all environment variables and credentials stored within the file to the shell session and any subsequent commands, representing a significant data exposure risk.
[COMMAND_EXECUTION]: The skill executes a local Node.js script scripts/worldlabs-generate.mjs to perform core logic. Because the script's source code is not included in the manifest, its operations on the local file system and network cannot be verified.
[EXTERNAL_DOWNLOADS]: The skill installs the @sparkjsdev/spark package from the npm registry and connects to the api.worldlabs.ai external domain to facilitate its 3D generation features.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection. \n
Ingestion points: User-provided prompt and image arguments passed to the CLI script. \n
Boundary markers: None; the instructions do not provide delimiters or safety warnings for processing untrusted external content. \n
Capability inventory: The skill has the ability to execute shell commands, access the network, and write to the local file system. \n
Sanitization: There is no evidence of validation or sanitization of user inputs before they are interpolated into command-line arguments. Additionally, there is a metadata discrepancy between the author 'OpusGameLabs' and the expected developer context 'M4cs'.

worldlabs