dune-mcp-skill
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the execution of the
uxccommand-line tool and creates a command aliasdune-mcp-clito interact with the Dune MCP server. - Evidence: SKILL.md and references/usage-patterns.md contain instructions to run
uxc auth,uxc link, and multiple operational commands such asdune-mcp-cli searchTablesanddune-mcp-cli executeQueryById. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the Dune API, which creates a surface for indirect prompt injection if the ingested data contains malicious instructions.
- Ingestion points: Untrusted data enters the agent context through API responses from
searchDocs,searchTables, andgetExecutionResultsas described in SKILL.md and usage-patterns.md. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' directives for data fetched from the external API.
- Capability inventory: The skill has the capability to execute shell commands and manage credentials via the
uxctool (SKILL.md). - Sanitization: No evidence of data sanitization, escaping, or validation of the API-returned content is provided before the agent processes it.
Audit Metadata