AILANG Roadmap

This page is automatically generated from the design_docs/planned directory. Each item represents a planned feature or improvement with a detailed design document.

For completed features, see Design Documents.

Planned for v1.1.0

• M-EXECUTOR-VARIANTS Sprint Plan
• M-EXECUTOR-VARIANTS: Per-Agent Docker Image Variants
• M-PROCESS-MODES: Replay-Contract Modes for the Process Effect

Planned for v1.0.0

• Global Collaboration Hub - Cross-Computer Agent Collaboration
• M-AGENT: std/agent Module for AI Agent Orchestration
• M-CSP-SESSION-TYPES: CSP Concurrency with Session Types
• M-DX-PACKAGE-DOGFOODING: DX Issues Found Building AILANG Packages in AILANG
• M-EFFECT-REFINEMENT: Parameterised Effects and Unified Replay Contracts
• M-ENTROPY: Semantic Entropy Budgets
• M-EU-COMPLIANCE-EFFECTS: Regulatory Trigger Effects for Agentic Systems
• M-PERF4: Bytecode Interpreter (Stretch Goal)
• M-QUASI: Typed Quasiquotes (String Templates)
• Sprint Plan: M-TYPE-V2-MIGRATION
• M-TYPE-V2-MIGRATION: Complete TFunc → TFunc2 Type System Migration

Planned for v0.17.0

• M-AGENT-LOOP-ARCHITECTURE: Where Should the Multi-Turn Agent Loop Live?
• M-AI-OPENAI-LOCAL-ENDPOINT-RELAX — Allow empty OPENAI_API_KEY when OPENAI_BASE_URL is custom
• M-AI-REASONING-EFFORT — Cross-Provider Request-Side Reasoning Control
• M-AI-STREAMING-HELPER Sprint Plan
• M-AI-STREAMING-HELPER: AI Token Streaming Helper + Cross-Domain Discovery
• M-BENCH-MOTOKO: Motoko Agent as a Benchmark Executor
• M-EXTERNAL-CONSUMER-DX Sprint Plan
• M-EXTERNAL-CONSUMER-DX: Diagnostics & Artifacts for External AILANG Projects

Planned for v0.16.0

• M-AGENT-SAFE-RUNNER: Turnkey Sandboxed Execution for AI-Authored AILANG Programs
• M-AI-EFFECT-MODES-FOLLOWUPS: Close the Loose Ends from v0.15.0 AI Modes

Planned for v0.15.2

• M-EVAL-FOLLOWUPS-V0151

Planned for v0.15.0

• M-AI-PROVIDER-CONFIG: Config-Driven AI Providers via Package Manifests
• M-BENCHMARK-SECTION: Multi-Page Benchmark Section for AILANG Website
• M-CASCADE-OBSERVABILITY: Surface Cascade AI Conversation History
• M-EVAL-GAP-FIXES: Eval-discovered AILANG and harness improvements
• M-EVAL-RESULTS-FOLDER-STRUCTURE: Model-First Eval Results Layout
• M-EVAL-TRUST-SIGNALS: External Trust Signals for AILANG Eval Results
• M-FEEDBACK-TRIAGE-GATE: Cost & abuse gate for the public feedback → agent pipeline
• Sprint Plan: M-PKG-FEEDBACK-LOOP (Test + activate per-package feedback + macOS notifier)
• M-PKG-FEEDBACK-LOOP: Validate the per-package feedback loop end-to-end

Planned for v0.13.0

• AILANG: Auto-Caps and Capability Inference
• M-REPL1: REPL Persistent Type Bindings & Module Loading
• M-TOOLING: Deterministic CLI for AI Agents (v0.3.15)
• M-ARCH-BOUNDARIES: Formalize Dashboard/Core Separation
• M-ARCH1: AI Provider Base Class
• M-ARCH2: Daemon God Object Refactor
• M-ARCH3: Task Classification Consolidation
• M-ARCH4: Executor Stream Processor
• M-ARCH5: Error Handling Strategy
• M-BUG-LETREC-SINGLE-CALL: Letrec with Single Recursive Call Fails
• M-BYTECODE-VM-PARITY-BUGS — Remaining VM/Eval Divergences
• M-CALL-SUGAR: Optional Parenthesized Call Syntax
• M-CLOUD-EVAL: Distributed Cloud Evaluation Workers
• M-CLOUD-OBSERVATORY: Observatory Firestore Backend for Cloud Run
• M-CONCURRENCY-LEVERAGE: Leverage Fork() Across All Execution Modes
• Sprint Plan: M-COORD-CODEX (Codex Executor for Coordinator + Evals)
• M-COORD-CODEX: Codex Executor for Coordinator + Evals
• M-COORD-THINKING: Extended Thinking Levels for Coordinator Skills
• M-COPILOT-CLI: GitHub Copilot CLI Integration
• M-CRYPTORAND: Cryptographic Randomness as a First-Class Effect
• M-D4: Design-Doc-Driven Development (D4)
• M-DASHBOARD-PUBSUB-EVENTS: Dashboard Pub/Sub Event Subscriber
• Dashboard Simplification - Remaining Work (v0.7.0)
• M-DX-AGENT-EVAL-GAPS: Agent DX Improvements from v0.9.0 Eval Analysis
• M-DX-AI-DISCOVERY: Improve AI Agent Stdlib Discovery
• M-DX-EXAMPLES-COVERAGE: Stdlib Examples for Every Module
• M-DX-EXPECTED-FAIL-FIXES: Fix Remaining Expected-Fail Examples
• M-DX-JSON-BOOL: JSON Boolean Coercion and Firestore Encoding Consistency
• M-DX-MATCH-HOF: Fix match Inside Block-Body Lambdas in HOF Arguments
• M-DX-RECORD-CONS: Record Literal + :: Cons Pattern Bug
• M-DX-SPLIT-ARG: Compile-Time Warning for Reversed split Arguments
• M-DX-TAPP-TRECORD: Type Inference Bug with Nested [[RecordType]]
• Inline Tests Documentation & Examples (M-DX23)
• M-DX26: Property Test "Empty Program" Bug
• M-DX27: GitHub Repository Search Fallback for ailang docs search
• M-ERROR-PROP: Error Propagation Operator (?)
• Sprint Plan: M-EVAL-BOUNDED-PIPELINE
• M-EVAL-BOUNDED-PIPELINE: Fused Bounded Combinators + Memory Ceiling
• M-EVAL-EXPAND: Expanding the AILANG Eval Bench (Harnesses, Languages, Open-Source Models)
• M-EVAL-LAZY-PIPELINE: Lazy List Pipelines to Fix OOM on Large Documents
• M-EXEC-HIERARCHY-REFACTOR: Executive Hierarchy Graph Visualization
• Sprint Plan: M-FILE-HANDLING
• M-FILE-HANDLING: fileData/fileUri Support & serve-api POST Param Fix
• M-GEMINI-INTERACTIONS: Google Gemini Interactions API Support
• Sprint Plan: M-GENERIC-PIPELINE
• Bug: show() builtin fails in serve-api when transitive deps also use show()
• M-LOCOBENCH: Long-Context Benchmark Integration
• Improve Module Loading Error Messages for Standalone/Scratch Files
• Sprint Plan: M-OBS-CONFIGURABLE-SPAN-FILTERING
• M-OBS-CONFIGURABLE-SPAN-FILTERING: Configurable Observatory Span Filtering
• M-PIPE-OPERATOR: Pipe operator and Option chaining
• M-PKG-TRUSTED-AUTONOMOUS-EVOLUTION: Secure Autonomous Package Evolution
• Sprint Plan: M-PROTOCOL-SUPPORT
• M-PROVENANCE: Code Provenance Tracing & Agent Trace Export
• M-REFLECT: Structural Reflection & User-Defined Type Classes
• M-TASK-GRAPH-SPANS-UNIFICATION: Fix TaskHierarchyGraph Filtering
• M-TYPEENV-SUB: TypeEnv Substitution Gap — ADT Return Types Lost in Cross-Module Exports
• M-UI-REFACTOR: Refactor UI Folder for AI-Friendly File Sizes
• Sprint Plan: M-UI-REFACTOR - AI-Friendly File Sizes
• Sit-Rep: Determinism vs. Unpredictability — The Rand Axiom Tension

Planned for docparse-billing

• DX Improvements Discovered During Billing Package Creation
• M-BILLING: DocParse Billing and Agent-Assisted Payment
• M-BILLING: DocParse Responsibility
• M-BILLING: ailang-multivac Responsibility
• M-BILLING: ailang-packages Responsibility

Long-term Vision

AILANG is designed as a deterministic language for autonomous AI code synthesis. The long-term roadmap includes:

• Structural Reflection - Typed quasiquotes and AST manipulation
• Schema Registry - Machine-readable type and effect definitions
• Capability Budgets - Resource-bounded effects
• Training Data Export - Execution traces for AI self-training

For the complete vision, see Why AILANG and Vision.

---

Generated at build time. 99 planned features across 8 upcoming versions.