Repository navigation
Roadmap
Quick-Scan Infobox
Aspect Details Target User Contributors planning involvement, educators anticipating classroom tools, developers tracking releases—familiarity with version control helpful. Core Goal Outline phased progress: From pedagogy-focused stability to scalable applications, with clear entry points for participation. Time to First Insight 5 minutes: Review current phase, scan metrics table. Key Analogy A curriculum sequence: Each phase builds on the last, with built-in checks (metrics) to ensure readiness before advancing. Evolution Note This document updates quarterly or with major PRs (label "roadmap-update"); v1.0 baseline as of November 2025. Known Limitation Timelines flexible based on contributions; community input shapes priorities.
This diagram maps the phases as a progression: Each builds stability while expanding reach, with signals (metrics) guiding adjustments.
journey
title "Phased Development: From Foundations to Emergence"
section Pedagogy Pack (v0.1.1)
Deliver Tools : 3 : Interactive Demos
Verify Signals : 2 : Drift and Slack Checks
Pilot Feedback : 4 : Educator Trials
section Enterprise Hardening (v0.1.2)
Deploy APIs : 4 : Load-Ready Services
Tune Constraints : 3 : Shadow Price Alerts
Scale Benchmarks : 3 : Multi-Domain Tests
section Multi-Domain (v0.2.0)
Adapt Metrics : 5 : Cross-Area Fitting
Add Plugins : 4 : Custom Observers
section Verification (v0.3.0+)
Prove Invariants : 5 : Lyapunov Theorems
Test Emergence : 4 : Agent Coordination
Caption: Journey through phases, with effort estimates per step. Alt text: Timeline diagram showing four phases with sub-tasks and durations, using light colors for readability.
This sequence ensures each phase verifies prior work: Pedagogy tools must show stable signals before hardening for scale.
Status: Imminent release (target: December 2025).
Vision: Equip educators with self-correcting tools for AI-assisted learning, emphasizing observable signals over black-box outputs.
Key Deliverables:
- Interactive REPL for live routing demos.
- Student portfolio generator from certificates.
- Constraint experiments module (e.g., vary shadow prices for "what-if" scenarios).
Success Metrics (Exit Criteria):
| Metric | Target | Verification |
|---|---|---|
| Certificate Interpretability | 90% of educators rate "clear" in pilot (n=20) | Survey after 1-week trial. |
| Signal Stability | Drift < -0.01 across 100 routes | Batch tests in Examples. |
| Integration Ease | <5 minutes to add to lesson plan | Timed setup in cookbook recipes. |
Community Callouts:
- PRs for education recipes: Focus on Montessori parallels.
- Pilot testers: Run demos with students; share feedback via issues labeled "pedagogy-pilot".
This phase closes the loop on early feedback: Tools now provide immediate, intrinsic checks, like a material that reveals mismatches without teacher intervention.
Status: Planning (Q1 2026).
Vision: Strengthen for production: Robust APIs, load handling, and ethical safeguards via shadow prices.
Key Focus Areas:
- Dockerized deployment with health checks.
- Dynamic constraint updates (e.g., runtime shadow price alerts).
- Multi-domain benchmarks (education, research, operations).
Success Metrics:
| Metric | Target | Verification |
|---|---|---|
| Throughput | 100 routes/min under load | Locust tests in production recipes. |
| Ethical Compliance | Shadow prices flag 95% of bias risks | Audit on diverse prompt sets. |
| Uptime | 99.5% in 24-hour stress test | Monitoring setup from cookbook. |
Community Engagement:
- Contributions: Harden API endpoints; label PRs "enterprise".
- Feedback: Simulate loads with provided scripts; report variances.
Here, hardening means extending bounds without brittleness—systems that adapt under pressure, much like a prepared environment that scales to group use.
These outline long-term directions, refined by proposals. Submit via issues (template below).
Vision: Extend to non-educational uses (e.g., research routing, operational triage).
Highlights:
- Cross-domain metric adaptation.
- Plugin system for custom observers.
Metrics: 80% accuracy lift on benchmarks like RouterBench.
Vision: Prove emergent behaviors (e.g., multi-agent coordination) with theorem checkers.
Highlights:
- Lyapunov proofs for pipelines.
- Entropy bounds in large-scale runs.
Metrics: Verified invariants in 10K routes; community-verified cases.
Development Philosophy: Bounded Progress
Phases release incrementally: Test one invariant (e.g., feasibility), advance only if met. This prevents overreach—each step verifiable, like checking tower stability before adding blocks.
Why Phased? Allows parallel contributions: Pedagogy in v0.1.1 doesn't block enterprise in v0.1.2.
Decisions evolve through input: Propose changes, discuss, implement.
Process:
- Open issue with proposal (use template).
- Label "roadmap-discuss"; tag stakeholders.
- Vote in weekly sync (thumbs-up on summary).
- Merge if consensus (>50% approval).
Proposal Template:
Title: [e.g., Add Multi-Agent Support to v0.2.0]
Summary: [1-paragraph overview]
Rationale: [Why now? Link to philosophy/metrics]
Deliverables: [Bullet list]
Metrics: [Table of targets/verification]
Risks: [Potential issues and mitigations]
Milestone Tracking:
- GitHub Projects: Board at [link].
- Syncs: Bi-weekly calls (calendar invite via #community).
Gotchas and Risk Management:
| Risk | Mitigation |
|---|---|
| Scope Creep | Strict metrics per phase; defer to future. |
| Dependency Delays | Parallel tracks (e.g., pedagogy independent of API). |
| Community Burnout | Optional involvement; credit in releases. |
Adaptive planning: If metrics slip, extend phase—progress over deadlines.
From single routes to coordinated systems: Tools that learn locally, stabilize globally. Success means educators with reliable aids, developers with verifiable APIs, researchers with provable bounds.
Get Involved:
- Educators: Test v0.1.1 demos; suggest lesson ties.
- Developers: PR hardening features; review APIs.
- Researchers: Propose verifications; benchmark domains.
- Enterprise Users: Simulate loads; report gaps.
Resources:
- Key Documents: Philosophy for vision; Cookbook for prototypes.
- External: GitHub issues for proposals.
Roadmaps as Living Documents: This evolves with use—review, propose, contribute.
Last Tuned: November 08, 2025 · Compitum v0.1.1 imminent · Open to Proposals