mistral-api

The mistral-api skill presents a coherent and proportionate API integration for Mistral models with function calling and code generation. There are no suspicious or unverifiable download/install patterns, and credentials are scoped to environment variables and standard Authorization headers. Data flows to an external API are expected given the purpose, and local tool invocation only occurs as part of legitimate function calling flows. Overall, the risk posture is low to moderate and aligned with its stated purpose, with standard mitigations (secure storage of API keys, validation of function tool outputs, and careful handling of streaming data).