The Agent Skills Directory

[SAFE]: The skill implements a robust security model for GraphQL supergraph configuration. It provides a dedicated validation checklist in validation/checklist.md that checks for common misconfigurations such as active introspection, enabled sandbox, or wildcard CORS in production environments.
[SAFE]: Secret management is handled correctly. The instructions and templates (e.g., templates/v2/production.yaml) consistently use shell environment variable expansion (${env.VAR}) for sensitive data like APOLLO_KEY, JWKS_URL, and JWT_ISSUER, preventing hardcoded credentials in configuration files.
[SAFE]: The skill uses version-aware logic to ensure configuration correctness between Apollo Router v1 and v2, reducing the risk of security regressions during migration.
[SAFE]: The included references and templates for Rhai scripts and coprocessors provide standard, safe implementation patterns for extending router functionality without introducing malicious execution vectors.
[SAFE]: External references and documentation URLs point to official apollographql.com domains, which are consistent with the skill's author metadata.

apollo-router