Skip to content

Repository files navigation

Recursive Meta-Cognitive Predictive Memory Kernel (RMCPMK)

RMCPMK is a synthetic cognitive substrate implemented in Rust. It is designed as a memory operating system (not a conventional database) with multi-tier memory hierarchy, predictive coding, causal-semantic reasoning, and metacognitive self-monitoring.

Workspace layout

  • crates/rmcpmk-core: kernel, cognitive modules, universal memory bus API, pluggable compressed storage engines.
  • crates/rmcpmk-api: Axum service exposing universal memory bus operations.
  • crates/rmcpmk-sim: simulation + benchmark-oriented prototype runner.
  • frontend: React + D3 testing UI for graph/metacognition/sleep/perf experimentation.
  • docs: architecture and retrieval diagrams.
  • benches: benchmark strategy and measurement plan.
  • scripts/final_build.sh: one-command final build pipeline.

Universal Memory Bus API

  • store()
  • recall()
  • associate()
  • predict()
  • reflect()
  • self_evaluate()
  • revise_belief()
  • estimate_uncertainty()
  • consolidate()
  • stream_context()

API endpoints

  • GET /health
  • GET /metrics
  • POST /store
  • POST /store_batch
  • POST /recall
  • POST /associate
  • POST /stream_context
  • POST /reflect
  • POST /predict
  • POST /estimate_uncertainty
  • POST /revise_belief
  • GET /self_evaluate
  • POST /consolidate

Runtime configuration

  • RMCPMK_STORAGE_DIR=/path/to/dir => use FileBackedCompressedEngine persistence.
  • RMCPMK_L0_CAPACITY=8192 => override working-memory ring size.
  • RMCPMK_REFLECTION_THRESHOLD=0.7 => adjust reflective retrieval trigger.
  • RMCPMK_HOST=0.0.0.0 + RMCPMK_PORT=8080 => bind address for API service.

Build and run

Quick start

cargo run -p rmcpmk-api
cargo run -p rmcpmk-sim

Frontend:

cd frontend
npm install
npm run dev

Final build (recommended)

./scripts/final_build.sh

Docker

docker compose up --build

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages