The Agent Skills Directory

[SAFE]: The skill is a standard implementation for chaos engineering automation. It interacts with the Harness MCP server to perform legitimate infrastructure testing tasks as described in its metadata.
[COMMAND_EXECUTION]: While the skill can initiate actions that delete pods or stress resources, these are the primary intended functions of a chaos engineering tool. These actions are performed through controlled MCP tool calls (e.g., harness_execute) rather than direct shell execution.
[DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or unauthorized network operations were detected. The skill requires standard identifiers like org_id and project_id to be provided by the user or environment at runtime.

chaos-experiment