Ariadne is a cross-service API dependency graph and semantic code navigation tool for microservice architectures. It exposes four core tools:

• query_chains — Search for cross-service API chains using a business term or endpoint name (e.g. "createOrder"). Returns ranked clusters of related GraphQL operations, HTTP endpoints, Kafka topics, and frontend queries across all indexed services.

• expand_node — Trace one-hop relationships from a known node (e.g. a Kafka topic, REST endpoint, or GraphQL mutation) to discover directly connected components with similarity scores.

• ariadne_help — Retrieve setup and usage guidance, including how to index microservices or troubleshoot empty results. Requires no database access.

• log_feedback — Record thumbs-up/thumbs-down feedback on results, stored locally in feedback.db to gradually improve cluster rankings over time.

It supports GraphQL, Spring (Java/Kotlin), TypeScript, and cube.js codebases; operates offline via static analysis; and integrates with MCP-compatible tools like Claude Code and Cursor.

Scans GraphQL SDL files to extract Query, Mutation, Subscription, and Type definitions for cross-service API dependency analysis and semantic code navigation.

Analyzes Kotlin code with Spring framework annotations to identify HTTP endpoints, Kafka topics, and backend client calls for microservice API dependency analysis.

Analyzes Spring framework code to identify HTTP endpoints via @RestController annotations, Kafka topics via @KafkaListener and application.yaml configurations, and backend client calls via RestClient/RestTemplate.

Scans TypeScript files to extract frontend GraphQL queries/mutations from gql template literals and REST API calls from axiosRequest/fetch statements for cross-service feature mapping.