LogicAIn

LogicAIn is a high-fidelity toolkit that implements the Codebase-as-Data architecture. It turns a Python codebase into a queryable data graph, decomposing every function and class into a SQL registry, mapping the dependencies between them, and exposing the entire structure through a single CLI tool—no manual SQL required.

---

What it does

Layer	File	Purpose
AST Decomposer	src/ast_decomposer.py	Parses Python files and extracts every function, class, import, and constant.
Graph Engine	src/graph_engine.py	Builds a networkx directed graph from the registry and answers dependency queries.
Sync Script	sync_codebase.py	Walks src/, decomposes each file, and writes the results to logic_registry.db.
CLI	logic.py	Single entry-point for all registry operations (no SQL needed).

---

Quick Start

1. Install dependencies

pip install networkx

2. Sync your codebase into the registry

python logic.py sync

Scans src/, decomposes every .py file, and upserts all logic nodes into logic_registry.db.

---

Commands

logic sync

python logic.py sync
# optional overrides:
python logic.py sync --src ./my_src --db /tmp/my_registry.db

Scans the source directory and updates the registry.

---

logic trace <node>

python logic.py trace ASTDecomposer

Prints all direct and transitive callers and callees for the named function or class.

Trace: ASTDecomposer
====================

Direct callers  (1 found):
  ↑ run_sync

Direct callees  (3 found):
  ↓ get_full_decomposition
  ...

---

logic check

python logic.py check

Reports:

• Circular dependencies — code that loops back on itself.
• Orphaned nodes — functions or classes that are not called by anything and do not call anything.

Returns exit code 1 if any circular dependencies are found.

---

logic context <node>

python logic.py context my_function

Outputs a Markdown block containing the node's metadata and its immediate dependencies. Designed for copy-pasting into an LLM prompt.

# Logic Context: `my_function`

## `my_function`
- **Type**: function
- **File**: `src/my_module.py`
- **Lines**: 42–58
- **Arguments**: `radius: float`
- **Returns**: `float`
- **Calls**: `helper_one`, `helper_two`
...

---

Environment variables

Variable	Default	Description
LOGIC_DB_PATH	logic_registry.db	Path to the SQLite registry database.

---

Database schema

The SQLite database (logic_registry.db) mirrors the target PostgreSQL schema:

CREATE TABLE logic_registry (
    id          INTEGER PRIMARY KEY AUTOINCREMENT,
    node_name   TEXT    NOT NULL,
    node_type   TEXT    NOT NULL,   -- 'function' | 'class'
    source_file TEXT    NOT NULL,
    line_start  INTEGER,
    line_end    INTEGER,
    docstring   TEXT,
    logic_calls TEXT,               -- JSON array of callee names
    metadata    TEXT,               -- JSON object (args, return_type, …)
    synced_at   TEXT,
    UNIQUE(node_name, source_file)
);

Switching to PostgreSQL requires only changing the connection call in sync_codebase.py and src/graph_engine.py.

---

Branch

All work on this toolkit lives on the codebase-as-data-v2 branch and is intentionally kept separate from the main TFusion project.