The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides an installation command curl -fsSL https://superset.sh/cli/install.sh | sh. This is a high-risk pattern that pipes a remote script directly into a shell, allowing for arbitrary code execution without giving the user or the system a chance to verify the script's integrity or content.
[COMMAND_EXECUTION]: The skill uses several shell commands to audit and manage local and remote environments, including superset auth whoami, superset workspaces create, and superset agents list. This grants the agent significant control over the user's workspace and environment configuration.
[EXTERNAL_DOWNLOADS]: The skill downloads a script from an external, non-whitelisted domain (superset.sh) to install system-level tooling.
[INDIRECT_PROMPT_INJECTION]: The skill processes output from command-line tools that could be influenced by external data, creating a risk of indirect prompt injection.
Ingestion points: Data is ingested from the outputs of superset agents list, superset automations list, and superset tasks list (SKILL.md).
Boundary markers: There are no explicit markers or instructions telling the agent to ignore potentially malicious instructions embedded in the tool outputs.
Capability inventory: The skill has the capability to execute system commands and create/modify resources like automations, workspaces, and agents (superset automations create, superset workspaces create, superset agents create).
Sanitization: The skill lacks any mentioned validation or sanitization steps for the data retrieved from the CLI tool audit.

superset-10x