The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references the installation of several Node.js packages from the official npm registry and the skill itself from a GitHub repository owned by the vendor.
Evidence: npm install @neo4j/graphql neo4j-driver graphql @apollo/server and npx skills add https://github.com/neo4j-contrib/neo4j-skills.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of a GraphQL API which represents a surface that processes untrusted external data (queries and mutations).
Ingestion points: The agent ingests schema definitions and configuration instructions from SKILL.md and references/search-directives.md.
Boundary markers: The skill utilizes standard Markdown code blocks to delimit instructions and highlights the use of @authentication and @authorization directives to define security boundaries.
Capability inventory: The skill uses Bash for dependency management and WebFetch to retrieve external documentation.
Sanitization: The documentation explicitly teaches the use of parameterized inputs ($param) and built-in authorization filters to sanitize data flow between the API and the Neo4j database.

neo4j-graphql-skill