The Agent Skills Directory

[SAFE]: The skill queries activity data from a user-specified LiteLLM proxy using curl. This is consistent with its stated purpose of viewing usage and spend metrics.
[SAFE]: Credentials such as the LITELLM_API_KEY are requested from the user and used as environment variables or arguments for authorization headers. The skill follows safe practices by not hardcoding secrets or attempting to exfiltrate them.
[SAFE]: The skill uses a local Python script to summarize and format API responses. This script processes JSON data from standard input and does not perform any dangerous dynamic code execution or remote script fetching.
[SAFE]: All external references point to the official LiteLLM documentation, which is maintained by the skill's author (BerriAI).

view-usage