About waccy

Intelligent Financial Modeling Platform for Small Businesses - financial accounting and managerial accounting for the AI era; financial operating models for businesses in Python

d

Published by

decisionnerd

Visit View Profile

README.md

View on GitHub

WACCY

Intelligent financial modelling for small businesses — built in Rust.

Extracts, maps, and synthesises financial data from QuickBooks Online and SEC EDGAR into institutional-quality three-statement models.

Install

# Homebrew (recommended)
brew tap DecisionNerd/tap && brew install waccy

# Cargo
cargo install --git https://github.com/DecisionNerd/waccy

# Python (query-only)
pip install waccy

# Claude Code / Cursor skills
npx skills add DecisionNerd/waccy

# Universal script (also registers MCP server)
curl -fsSL https://raw.githubusercontent.com/DecisionNerd/waccy/main/scripts/install.sh | bash

Quick start

# Extract financial data
waccy extract quickbooks
waccy extract edgar --option cik=0001234567

# Build three-statement model
waccy model

# Query results
waccy query "SELECT * FROM records WHERE statement_kind = 'income_statement'"

MCP server

Register the MCP server (waccy-mcp) with your AI client:

# Claude Code
claude mcp add waccy $(which waccy-mcp)

Then ask your AI agent: "Build a three-statement model from my QuickBooks data."

Distribution surfaces

Surface	Package	How
CLI	`waccy` binary	Homebrew / `cargo install`
MCP server	`waccy-mcp` binary	Registered via `claude mcp add`
Python	`waccy` on PyPI	`pip install waccy`
Agent skills	11 skills	`npx skills add DecisionNerd/waccy`

Architecture

crates/
  waccy-core/     # Shared library: models, extraction, classification, modeling, query
  waccy-cli/      # Binary: waccy CLI
  waccy-mcp/      # Binary: waccy-mcp MCP server (JSON-RPC 2.0 over stdio)
  waccy-python/   # PyO3 cdylib: Python bindings (built with maturin)
legacy/           # Original Python implementation (archived)

Development

cargo build --release          # Build binaries
cargo test --all               # Run tests
cargo clippy --all-targets     # Lint
cargo fmt                      # Format

# Python wheel
cd crates/waccy-python && maturin develop

# Full setup
just setup

License

MIT

waccy