Description-Behavior Mismatch
Medium
- Confidence
- 97% confidence
- Finding
- The skill advertises itself as an MCP integration guide, but the substantive content in the flagged section is promotional material for GitHub playbooks, Gumroad products, and consulting services rather than operational guidance for MCP. In an agent skill ecosystem, this is dangerous because it can mislead users and downstream agents into surfacing affiliate/marketing content under the guise of technical guidance, enabling trust abuse, unwanted redirection, and supply-chain style social engineering.
