ValkyrAI: Your AI-Powered Workflow Companion
ValkyrAI v0.6.0 BETA is the API-native automation hub that backs the Valhalla Suite. Built on ThorAPI and Spring Boot 3.4, it translates OpenAPI specs and workflow definitions into deployable services, TypeScript clients, dashboards, and Model Context Protocol tools that ValorIDE can invoke on demand.
Platform Pillars
- Code generation-first – convert schemas into secure CRUD services, React clients, and documentation with consistent naming, validation, and RBAC baked in.
- Agentic workflow runtime – orchestrate tasks, execution modules, and long-running jobs with audit-friendly lifecycle APIs.
- Security everywhere – SecureField encryption, system user impersonation helpers, and policy-driven sharing aligned with enterprise RBAC requirements.
- Model-aware integrations – bundled connectors for Anthropic, OpenAI, Google Gemini/Vertex, AWS Bedrock, and ValkyrAI-managed models expose consistent inference primitives to downstream tools.
- Operational visibility – Gridheim dashboards, enhanced 6D reporting, and token/cost telemetry give teams a real-time picture of system health.
Architecture Snapshot
- Parent POM (
valhalla) – coordinates a multi-module Maven build targeting Java 17. thorapi/– generators, DTOs, and SecureField primitives consumed by downstream modules and ValorIDE.valkyrai/– primary Spring Boot service (controllers, services, MCP endpoints, AWS helpers, WebSocket brokers).gridheim/– React dashboards for operations, cost management, and workflow monitoring.web/– Docusaurus + Vite TypeScript projects that render this documentation and valkyrlabs.com marketing collateral.
ValkyrAI & ValorIDE
ValorIDE uses ValkyrAI as the default backend for ThorAPI project generation, workflow execution, billing telemetry, and MCP orchestration. When a developer kicks off a ThorAPI build or requests a workflow rollout from the IDE, the call flows through the valkyrai service, relies on artifacts in thorapi/, and surfaces progress back through the ValorIDE webview.
Feature Highlights (2025)
- Workflow CRUD, execution module lifecycle endpoints, and task queues.
- SecureField KMS integrators (AWS KMS, GCP KMS, vault-friendly abstractions).
- Built-in MCP servers for operations like project scaffolding, dataset ingestion, and report generation.
- Chromatic-tested web dashboards and Cypress integration suites for UI flows inside
web/typescript/. - Docker + Maven profiles for containerized builds and CI pipelines.
Where to Go Next
- Dive deeper with the ValkyrAI Documentation index.
- Follow the Getting Started guide to compile and launch the stack locally.
- Explore the Workflow Engine to understand how execution modules plug into your processes.