The Agent Skills Directory

[Data Exposure & Exfiltration] (SAFE): The skill uses safe secret placeholders like ${ secrets:OPENAI_API_KEY } instead of hardcoding credentials. No sensitive file paths or unauthorized network operations were detected.
[Prompt Injection] (SAFE): No instructions attempting to override agent behavior, bypass safety filters, or extract system prompts were found.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not install external packages or execute remote scripts. The provided curl example is a standard API query template targeting a local endpoint.
[Indirect Prompt Injection] (SAFE): While the skill facilitates the configuration of system prompts, it is a static documentation resource and does not ingest untrusted external data within the context of the skill itself.

spice-models