cadence

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to connect to a remote MCP server at https://cadence-mcp.up.railway.app/mcp and links to documentation on cadence-lang.org. These resources are official components of the Cadence development environment provided by the author.
  • [COMMAND_EXECUTION]: The documentation includes setup steps using npx to install the skill (npx skills add outblock/cadence-lang.org) and configure an MCP server using mcp-remote. These commands are standard for the integration of the vendor's provided development tools.
  • [PROMPT_INJECTION]: The skill utilizes an MCP server that ingests documentation data from external sources. This identifies a surface for indirect prompt injection, though it is limited to technical lookup and diagnostic functions.
  • Ingestion points: MCP tools like search_docs, get_doc, and browse_docs fetch content from the remote server at cadence-mcp.up.railway.app.
  • Boundary markers: No explicit markers or warnings are specified for the tool outputs in the skill body.
  • Capability inventory: The skill provides access to LSP tools such as cadence_check, cadence_hover, cadence_definition, and cadence_symbols.
  • Sanitization: Content is treated as technical reference data for smart contract auditing and code checking.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 05:58 PM