explore-metric
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references an MCP server configuration at
https://mcp.confidence.dev/mcp/flags. This is an official domain for Spotify's Confidence platform, and the reference is part of the standard setup for the skill's dependencies. - [COMMAND_EXECUTION]: Provides instructions for the user to manually execute a command to install the necessary MCP extension. This is a one-time setup step and is not performed automatically by the skill's internal logic.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied event and fact table names to generate URLs and perform lookups.
- Ingestion points: User input via arguments and interactive questions.
- Boundary markers: Not explicitly defined in the prompt instructions, though the final output is a structured URL.
- Capability inventory: The skill uses read-only MCP tools to list tables and event definitions, and generates a navigation URL.
- Sanitization: The instructions require the agent to perform URL-encoding on all resource names to prevent malformed links.
Audit Metadata