constructive-graphql-query

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches runtime schema metadata from external GraphQL endpoints (e.g., fetch('/graphql' with SCHEMA_INTROSPECTION_QUERY and the PostGraphile _meta query shown in SKILL.md and references/meta-introspection.md), and then uses that untrusted introspection/_meta data to build query and mutation names and selections that directly drive subsequent tool behavior.