The Agent Skills Directory

[SAFE]: No security issues detected. The skill's primary functionality is to send user-provided design data to a remote API for analysis.
[COMMAND_EXECUTION]: The script scripts/run.py is provided for the user to execute locally to interact with the API. It uses standard Python libraries for networking and parameter handling.
[EXTERNAL_DOWNLOADS]: The skill communicates with the vendor's domain ai-skills.ai to execute its logic. It also references external open-source material from skills.sh, which is presented neutrally as the source of the skill's logic.
[PROMPT_INJECTION]: The skill has an indirect prompt injection attack surface as it processes external content provided by the user.
Ingestion points: Data enters the system via materialText, materialUrl, and materialFile fields defined in references/form-schema.json.
Boundary markers: None are explicitly specified in the local file instructions.
Capability inventory: The skill uses network communication to send data to the provider's API via scripts/run.py.
Sanitization: There is no evidence of input validation or sanitization in the provided scripts.

ai-stitch-design-taste