# Roadmap
This document outlines the planned direction and priorities for DevOps AI Toolkit.
This roadmap is a living document and may change based on community feedback and project priorities.
## Short-term (Next 1-3 Months)
- Modular plugin architecture - transform dot-ai into extensible orchestrator (PRD #342)
- AI provider comparison and benchmarking infrastructure (PRD #139)
- Guided setup system to simplify onboarding (PRD #165)
- Standardized embedding architecture on Vercel AI SDK (PRD #176)
- Deployment documentation & example-based learning (PRD #228)
- Git operations for recommend tool - push generated manifests to Git repos (PRD #362)
- Argo CD support for operate tool - detect managed resources and modify source repos (PRD #363)
- Documentation validation system - crawling, readability, syntax validation (PRD #262)
- Consolidate duplicated constants and messages to eliminate code duplication (PRD #323)
- Update to Kimi K2.5 model support (PRD #353)
- Update OpenAI model to GPT-5.3-Codex when API available (PRD #369)
- User authentication & access control for multi-user MCP deployments (PRD #360)
- User-specific permissions - align AI view with user's Kubernetes RBAC (PRD #361)
## Medium-term (3-6 Months)
- Auto-generated Go CLI as alternative to MCP - self-contained multi-arch binaries from OpenAPI spec (PRD #371)
- MCP server integration for extended tool capabilities - observability, tracing, logging (PRD #358)
- Policy extraction from knowledge base with human-in-the-loop review (PRD #357)
- MCP tool filtering - server-side allow/deny lists to control tool exposure (PRD #347)
- Elicitation enhancement across MCP tools - systematic review and improvement of confirmation, clarification, and adaptive questioning (PRD #307)
- Skills distribution via MCP - install-skills prompt and enhanced skill format (PRD #290)
- Multi-step workflow distributed tracing (PRD #197)
- Integrate patterns & policies into remediation tool (PRD #227)
- Pattern & policy learning system with AI-driven suggestions (PRD #218)
## Long-term (6-12 Months)
- Agent-to-Agent (A2A) protocol support for multi-agent orchestration (PRD #199)
## How to Contribute
We welcome contributions toward any roadmap items:
1. Check the [issue tracker](https://github.com/vfarcic/dot-ai/issues) for related work
2. Comment on issues or open discussions
3. See [CONTRIBUTING.md](../CONTRIBUTING.md) for guidelines
---
*This roadmap reflects current plans but is subject to change.*
