12 projects
bezalel-engine
Bezalel — a self-extending reasoning engine: sandboxed primitives plus a human-gated self-extension loop.
mahalath
Self-sustaining multi-agent ontology builder with REM-style refinement.
ed-witness
Ed — the Witness Translation Engine: AI-assisted Bible translation reasoning with inspectable multi-LLM debate.
deborah
Deborah — Cairn process language, cognitive contracts, thin PLAN interpreter, estate hooks, and substrate slice.
milcah
Milcah — the Coherence Engine: recursive, multi-LLM coherence pressure-testing of frameworks and worldviews.
galeed
Galeed — the cross-project trace/log spine: a structured event stream separating process telemetry from final output.
keturah
Keturah — a uniform, MCP-bridgeable manifest of a product's LLM-consumable interfaces.
hanani
Hanani — the Intelligence Synthesis Engine: multi-source geopolitical evidence synthesis with inspectable reasoning.
tirzah
Local graph-based memory and retrieval layer for LLM interactions.
hoglah
Lightweight local-first job queue manager and Ollama wrapper for resource-constrained environments.
cairn-lang
Compatibility shim — `cairn` split into `deborah` (process language) and `huldah` (human-systems analysis).
huldah
Huldah — human-systems analysis over Deborah process descriptions: human factors, UI evidence, layout load, live observation, and interface recommendations.
