superset

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill documents an installation method using a shell script fetched from https://superset.sh/cli/install.sh and piped to the shell (| sh). This URL belongs to the vendor's primary domain and represents a standard distribution mechanism for the CLI tool.
  • [COMMAND_EXECUTION]: The skill facilitates the execution of shell commands and the spawning of agents within isolated workspaces (e.g., superset workspaces create --command, superset terminals create). These are core features of the orchestration tool described in the instructions.
  • [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes user-provided prompts and commands.
  • Ingestion points: Command line arguments --prompt and --command used in various CLI commands within SKILL.md.
  • Boundary markers: None identified in the command syntax provided.
  • Capability inventory: Shell process execution, agent lifecycle management, and automation scheduling.
  • Sanitization: Not explicitly documented; the skill relies on the underlying CLI implementation for input handling.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 05:51 AM
Security Audit — agent-trust-hub — superset