Making AI systems transparent and reliable through open infrastructure
Navigation: This is the repository overview.
- Start Here — The single front door to SIL
reveal - Progressive code exploration that reduces AI agent token usage by 86%:
pip install reveal-cli
reveal your_code.pySee structure first (50 tokens) instead of reading entire files (500+ tokens). Progressive disclosure applied to code.
More production tools → - morphogen, tiacad, genesisgraph, beth, tia
The Semantic Infrastructure Lab builds open infrastructure that makes intelligent systems inspectable and reliable.
Contemporary AI systems are powerful but lack:
- Explicit meaning - Stable, inspectable semantic structures
- Persistent memory - Durable continuity across tasks
- Inspectable reasoning - Traceable chains of inference
- Cross-domain composition - Audio, CAD, simulation working together
- Verifiable provenance - Clear lineage of transformations
SIL builds this missing layer: the Semantic Operating System - a 6-layer architecture providing semantic memory, universal representations, deterministic execution, multi-agent coordination, and transparent interfaces.
Status: 12 projects, 4 in production, 3,100+ tests, used daily in real workflows.
→ Complete Documentation - Philosophy, architecture, research, and guides
Quick links:
- Founder's Letter (5 min) - Why SIL exists
- Manifesto (10 min) - You Only Live Once - Why we build
- FAQ - Common questions answered
- Unified Architecture Guide (30 min) - How it all fits together
→ Production Tools - Working systems you can use today
→ GitHub - How to get involved
Layer 5: Human Interfaces reveal, browserbridge
Layer 4: Deterministic Engines morphogen, riffstack
Layer 3: Agent Orchestration agent-ether
Layer 2: Domain Modules morphogen, tiacad, riffstack, sup
Layer 1: Universal Semantic IR pantheon
Layer 0: Semantic Memory semantic-memory
Cross-Cutting: Provenance genesisgraph, prism
Read the complete architecture →
- Code: Apache 2.0 - Use commercially, fork, modify
- Documentation: CC BY 4.0 - Share freely with attribution
- Email: scott@semanticinfrastructurelab.org
- GitHub: github.com/semantic-infrastructure-lab
- Website: semanticinfrastructurelab.org
Make meaning explicit. Make reasoning traceable. Build structures that last.