plugins
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
supersetCLI tool to perform plugin management tasks and call MCP tools. Execution is scoped to thesupersetbinary subcommands through platform configuration provided in the frontmatter. - [EXTERNAL_DOWNLOADS]: The skill allows the agent to install plugins from marketplaces, including third-party GitHub repositories via
superset plugins marketplace add. This process downloads and materializes new instructions (skills) into the local environment. These downloads are associated with the primary purpose of the skill as a plugin manager. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external plugin servers, including tool names, argument schemas, and the results of tool calls. This data is used to inform the agent's subsequent actions and tool selection.
- Ingestion points: Manifest data retrieved from plugin servers (SKILL.md) and tool output data from
superset mcp call-tool. - Boundary markers: None explicitly defined for tool outputs in the provided instructions.
- Capability inventory: Access to
supersetCLI commands for file modification (installation) and network-proxied tool execution. - Sanitization: No specific sanitization or filtering of external tool content is described, though instructions emphasize listing tools before execution.
Audit Metadata