Roadmap
LibreFang v0.1.0 release roadmap, organized into 4 sprint phases.
Current progress: 18/18 features complete, 1751 tests passing. Only install script domain deployment remains.
Overview
| Phase | Goal | Progress |
|---|---|---|
| Sprint 1 | Core Stability | 4/4 Complete |
| Sprint 2 | Competitive Parity | 4/5 Complete |
| Sprint 3 | Differentiation | 5/5 Complete |
| Sprint 4 | Polish & Launch | 5/5 Complete |
Sprint 1 — Core Stability
Addressing critical issues that affect basic usability.
| Feature | Description | Status |
|---|---|---|
| Token Context Management | Smart token budget control, reduced from ~45K to ~15-20K with auto-compaction and context window protection | Complete |
| Branding Assets | Full brand integration for desktop and web UI including icons, logo, and favicon | Complete |
| App Signing | Tauri desktop Ed25519 signing for auto-update verification | Complete |
| First-Run Experience | 8 wizard API endpoints fully implemented, 6-step onboarding flow ready out of the box | Complete |
Sprint 2 — Competitive Parity
Closing feature gaps with comparable solutions.
| Feature | Description | Status |
|---|---|---|
| Browser Screenshots | Browser automation screenshots displayed inline in chat | Complete |
| Chat Search | Ctrl+F shortcut, real-time filtering, and match highlighting | Complete |
| Skill Marketplace | 4 tabs (Installed, Community, MCP Servers, Quick Start) with search and categories | Complete |
| Install Script | One-line curl / irm installation, requires domain deployment | Pending |
| Setup Wizard | 6-step onboarding (Welcome → Provider → Agent → Try It → Channel → Done) with API key auto-detection | Complete |
Sprint 3 — Differentiation
Core innovations that go beyond competing solutions.
| Feature | Description | Status |
|---|---|---|
| Voice Interaction | Hold-to-record microphone, auto-transcription, TTS audio playback | Complete |
| Canvas Rendering | Secure sandboxed iframe rendering for web and desktop | Complete |
| Multi-Language SDKs | JavaScript (@librefang/sdk) and Python SDK with full API coverage | Complete |
| Observability | Prometheus-format metrics export (/api/metrics) with Grafana integration | Complete |
| Visual Workflows | Drag-and-drop DAG builder with 7 node types, bezier connections, and TOML export | Complete |
Sprint 4 — Polish & Launch
Performance tuning and release preparation.
| Feature | Description | Status |
|---|---|---|
| Multi-Session | Multiple independent sessions per agent with UI dropdown switcher | Complete |
| Config Hot-Reload | Auto-detect config.toml changes and apply without restart | Complete |
| Documentation | Comprehensive CHANGELOG and README updates | Complete |
| Load Testing | 7 benchmarks, API p99 latency < 5ms, throughput 2,792 req/s | Complete |
| Release Ready | All code complete, production audit passed | Complete |
Performance Benchmarks
| Metric | Result |
|---|---|
| Health endpoint p99 | 0.8ms |
| Agent list p99 | 0.5ms |
| Metrics throughput | 2,792 req/s |
| Concurrent reads | 1,728 req/s |
| Agent spawns | 97/s |
Feature Comparison
LibreFang compared against similar open-source solutions.
| Feature | Comparable Solutions | LibreFang | Advantage |
|---|---|---|---|
| Language / Performance | Node.js (~200MB) | Rust (~30MB single binary) | LibreFang |
| Channel Integrations | ~15 | 40 | LibreFang |
| Built-in Tools | ~19 | 41 | LibreFang |
| Security Systems | Token + Sandbox | 16-layer defense-in-depth | LibreFang |
| Agent Templates | Manual config | 30 pre-configured templates | LibreFang |
| Autonomous Capabilities | None | 7 Hands packages | LibreFang |
| Workflow Engine | Cron + Webhooks | Full DAG (parallel / loops) | LibreFang |
| Knowledge Graph | Flat vector store | Entity-relation graph | LibreFang |
| P2P Networking | None | OFP protocol | LibreFang |
| WASM Sandbox | Docker only | Dual-metered WASM | LibreFang |
| Desktop App | Electron (~200MB) | Tauri (~30MB) | LibreFang |
| Migration Tools | None | migrate --from openfang/openclaw | LibreFang |
| Skill Ecosystem | 54 bundled | 60 bundled | LibreFang |
| LLM Providers | ~15 | 27 providers / 130+ models | LibreFang |
| SDKs | TypeScript | JS + Python | Tie |
| Native Mobile | iOS + Android + macOS | Web responsive | Comparable |
| Voice Mode | Wake word + TTS | Mic + TTS playback | Comparable (slight) |
| Visual Workflows | None | Drag-and-drop builder | LibreFang |
LibreFang wins 15 out of 18 categories. Remaining gaps: native mobile apps and voice wake word.