New:Introducing Socket Scanning for VS Code Marketplace Extensions.Learn more →
Get Started
a

andreassigloch

npm

2.87K

total weekly downloads

Packages

Current Co-maintainers

Former Co-maintainers

13 packages

mcp-propstack

MCP server for PropStack real estate API integration

moneyflow-mcp

MCP server for moneyflow – query money flows, simulate policy changes, compare cohorts. Connects to money-flow.org.

@sigloch/contracts

Zod contracts for the SE model layer — ontology, rules and harness schemas (single source of truth)

@sigloch/graph-api-core

Framework-agnostic graph engine core — ontology-typed nodes/traces, rule evaluation, views, and the embedded Kuzu store (./kuzu)

@sigloch/graphcode

Governed graph substrate for coding agents (Claude Code, OpenCode, any MCP-stdio host) — Apply-Gate + on-disk Kuzu store + MCP tools

@sigloch/graphcode-client

Read-side client for a running @sigloch/graphcode host — host-socket calls, dashboard panels, readiness projection and the markdown view catalog. Deliberately free of Kuzu and the MCP SDK so a viewer can depend on it without pulling an embedded graph data

@sigloch/graph-cypher-wasm

Embedded Kuzu (WASM) adapter — Cypher execution and schema DDL for the graph engine

@sigloch/graphify

Extraction/Slicer — chaos documents → governed SE-graph candidate slices (slicer → auto-typing → suggestion-gate)

@sigloch/graph-view-edit

React/Vite viewer+editor over @sigloch/graphcode — 12 graph-views + 16 doc-views via a declarative View-Registry, plus an SE-Dashboard sibling route. Reads docs/graph/<member>.graph.json client-side (no 2nd Kuzu handle).

@sigloch/learning-core

Trajectory/outcome interfaces and math utilities for learning over agent runs

@sigloch/se-engine

SE analysis over a governed graph — metric vector, layer projection, rule apply/suggest, readiness computation

@sigloch/se-optimizer

Deterministic architecture analysis & optimization over an SE ontology graph — R6 topology metrics with layer projection and node weight; no LLM, no store, no transport

@sigloch/se-steering

Deterministic next-step steering over an SE readiness report — no LLM, no learning