This roadmap describes what Boruna is working toward. It is realistic, not aspirational marketing. Items without a milestone are under consideration but not scheduled.
The core execution engine is complete and tested. The goal for 0.1.x patch releases is stability and polish — no new features, only bug fixes and documentation.
What exists today:
.axlanguage: types, pattern matching, records, enums, capability annotations- VM: bytecode execution, capability enforcement, actor system, replay engine
- Workflow engine: DAG execution, topological sort, step isolation, approval gates
- Evidence bundles: hash-chained audit logs,
evidence verify/inspect - CLI: compile, run, trace, replay, workflow, evidence, lang, template, framework
- MCP server: 10 tools for AI coding agent integration
- Standard libraries: 11 pure-functional stdlib modules
- HTTP handler: real network calls (feature-gated, SSRF-protected)
- Package system: manifest format, dependency resolution, SHA-256 integrity
- 557+ tests covering all crates
Target: Q2 2026
Focus: making Boruna usable without friction for teams evaluating it.
-
boruna newcommand — scaffold a new workflow from a template interactively - Improved error messages — actionable diagnostics with suggested fixes for all common mistakes
-
boruna fmt— auto-formatter for.axfiles - Watch mode —
boruna run --watchre-runs on file change - Better
lang repair— handle more repair cases automatically - Evidence bundle diff — compare two runs side-by-side
- Workflow step output piping — pass step outputs as typed inputs to downstream steps
- Expanded stdlib —
std-llm,std-jsonlibraries
Target: Q3 2026
Focus: workflows that survive process restarts and handle long-running steps.
- Persistent workflow state — checkpoint and resume across process restarts
- Async step execution — steps that wait for external events (webhooks, approvals)
- Scheduled workflows — trigger workflows on a cron schedule
- Workflow versioning — run a workflow at a specific commit/version
- Step retry policies — configurable retry with backoff on transient failures
- Timeout enforcement — per-step and per-workflow execution limits
Target: Q4 2026
Focus: running Boruna in team environments.
- Distributed step execution — run steps on separate worker processes
- Workflow dashboard — web UI for run history, step status, evidence inspection
- Metrics and observability — Prometheus-compatible metrics endpoint
- Policy management — define and version capability policies as code
- LLM provider registry — configure and route between model providers
- Multi-environment support — dev/staging/production policy separation
Target: 2027
Milestone: the stable API surface is locked. 0.x programs compile and run unchanged.
- Stable
.axlanguage specification (versioned) - Stable workflow DAG schema (versioned)
- Stable evidence bundle format (versioned)
- Migration tooling for any pre-1.0 breaking changes
- Long-term support commitment for 1.x
- Security audit of the VM and capability enforcement
- Performance benchmarks and published baseline
These items are on the long-term radar but not scheduled:
- Commercial platform: hosted workflow execution, managed evidence storage, SSO, RBAC, compliance reporting — built on the open source core
- IDE integration: language server (LSP) for
.axsyntax, completion, and diagnostics in VS Code / Neovim - Model evaluation framework: run the same workflow against multiple LLM providers and compare evidence bundles
- Compliance templates: pre-built workflow patterns for common regulated use cases (SOC 2, HIPAA, financial audit)
- Cross-language FFI: call into Rust/Python libraries from
.axthrough a typed capability interface
Boruna will not become:
- A general-purpose programming language (use Rust, Python, etc. for that)
- An LLM framework (use LangChain, LCEL, etc. for that)
- A cloud provider (Boruna runs where you deploy it)
- A no-code tool (Boruna is for engineers)
See also: Stability, Limitations