The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the ax CLI to interact with the Arize platform for experiment CRUD operations, data export, and annotation.- [DYNAMIC_EXECUTION]: The skill facilitates the execution of local Python code through the ax experiments run --task task.py command and provides templates for custom inference scripts (infer.py). This is a core feature for running custom evaluation logic against datasets.- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from dataset exports and model outputs. If this data were processed as natural language instructions by an agent without proper boundaries, it could serve as a vector for indirect prompt injection.
Ingestion points: Data from ax datasets export output, runs.json, and annotations.json files.
Boundary markers: No explicit delimiters or instructions are specified for isolating the ingested data within the agent's prompt context.
Capability inventory: The skill allows for command execution via the ax CLI, local Python script execution, and file system write operations.
Sanitization: The instructions do not define specific sanitization or filtering logic for the content of the datasets or model outputs.

arize-experiment