The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill relies on npx -y @talesofai/neta-skills@latest for its operational commands. This pattern executes remote code from the public NPM registry at runtime without version pinning, which poses a supply chain risk.
[EXTERNAL_DOWNLOADS]: The skill triggers downloads of the @talesofai/neta-skills package from the NPM registry during tool invocation.
[DYNAMIC_EXECUTION]: The npx mechanism performs dynamic execution of a remote script upon every call to the creative tools.
[COMMAND_EXECUTION]: The skill executes shell commands to interact with the Neta API for content creation, character searches, and premium subscription features.
[DATA_EXFILTRATION]: The upload tool allows the agent to access local files via a user-controllable path and upload them to a remote service. This capability could be abused to exfiltrate sensitive files, such as environment variables or private keys, if the agent is manipulated via prompt injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes external character and collection data (e.g., via read_collection) and incorporates it into generation prompts without sufficient isolation or sanitization.
Ingestion points: External collection data, character search results, and local/remote media files processed via upload.
Boundary markers: None defined to isolate ingested content from system instructions.
Capability inventory: File reading, network transmission, and financial transaction initiation.
Sanitization: No sanitization of external input is required or performed by the instructions.

neta-creative