Skip to content

Releases: cjsthecoder/project_syx

Added Support for Multiple Models

Choose a tag to compare

@cjsthecoder cjsthecoder released this 18 Jun 20:15
b656e63

Added

  • Provider-agnostic LLM factory boundary with app code depending on shared LLM
    interfaces instead of concrete provider implementations.
  • App-owned provider/model registry for coherent OpenAI and Anthropic runtime
    model sets across chat, helper, builder, tagger, and Dream roles.
  • Anthropic provider support behind the existing LLM factory and runtime
    routing boundary.

Changed

  • Local setup and model-selection documentation now favor registry-driven
    provider defaults over hard-coded model inventory in environment files.
  • RAG route score thresholds are tuned to keep retrieval useful while filtering
    weaker candidates.

Syx v0.1.0 Public Preview

Choose a tag to compare

@cjsthecoder cjsthecoder released this 12 Jun 19:14

Syx v0.1.0 Public Preview

Initial public preview of Syx, an experimental local AI memory system for long-running projects, RAG-backed chat, sleep-cycle consolidation, dream-cycle synthesis, markdown memory artifacts, and read-only coding-agent memory access.

Included

  • Local project-based chat and memory.
  • React + Vite frontend served by FastAPI.
  • Provider-selectable LLM and embedding configuration.
  • File upload and FAISS-backed RAG for .txt and .md files.
  • Daily memory rolloff.
  • Sleep-cycle consolidation.
  • Experimental Dream-cycle synthesis.
  • Markdown memory artifacts with Syx entry boundaries and metadata.
  • Read-only local agent memory search endpoint and CLI tooling.
  • Docker and Docker Compose support.
  • Backend and frontend tests, CI, linting, formatting, and coverage guidance.

Not included

  • Production hardening.
  • Public internet deployment support.
  • Multi-user authentication or authorization.
  • Controlled external agent write-back.
  • MCP wrapper.
  • Hosted service support.

Notes

Syx is a research prototype intended for local or trusted environments only. Memory artifacts, uploaded files, chat history, logs, and agent query results may contain sensitive information.