​
Arc Documentation

Beta Release: Arc Memory SDK is currently in beta. We’re actively gathering feedback and improving the platform.
Arc is the local-first “what-if” engine for software: it captures the why behind every line of code, then simulates how new changes will ripple through your system before you hit Merge.

Quick Start

Get up and running with Arc in minutes

API Reference

Explore Arc’s programmatic interfaces

CLI Commands

Learn about the command-line interface

Examples

See real-world usage examples

​
What Arc Actually Does

  1. Record the why. Arc’s Temporal Knowledge Graph ingests commits, PRs, issues, and ADRs to preserve architectural intent and decision history—entirely on your machine.
  2. Model the system. From that history Arc derives a causal graph of services, data flows, and constraints—a lightweight world-model that stays in sync with the codebase.
  3. Predict the blast-radius. A one-line CLI (arc sim) spins up an isolated sandbox, injects targeted chaos (network latency, CPU stress, etc.), and returns a risk score plus human-readable explanation for the current diff.
  4. Prove it. Every simulation writes a signed attestation that links input code, fault manifest, and metrics—auditable evidence that the change was tested under realistic failure modes.

​
Why It Matters

Catch outages before they exist

Shift chaos left from staging to the developer’s laptop; trim MTTR and stop bad PRs at the gate.

Trust AI suggestions

Arc doesn’t just comment on code—it proves why a suggestion is safe (or isn’t) with sandbox data and a verifiable chain of custody.

Local-first, privacy-first

All graphs and simulations run inside a disposable E2B sandbox; no proprietary code leaves your environment.

Built to extend

The same graph and attestation layer will power live-telemetry world-models and multi-agent change control as teams grow.
Arc = memory + simulation + proof—your safety net for the era of autonomous code.

​
Installation

pip install arc-memory

​
Quick Start

1

Authenticate with GitHub

arc auth gh
This sets up your GitHub credentials for accessing repository data.
2

Build the knowledge graph

arc build
This analyzes your repository and builds a complete knowledge graph.
For faster updates after the initial build, use the incremental option:
arc build --incremental
3

Verify the graph

arc doctor
This checks the status and integrity of your knowledge graph.
4

Understand the why behind your code

arc why file path/to/file.py 42
This shows you the decision trail for line 42 in file.py, including commits, PRs, and issues.
5

Simulate the impact of changes

arc sim
This analyzes your latest commit, runs a simulation, and provides a risk assessment.

Explore More

See the full quick start guide for more detailed instructions

​
Documentation Sections

​
Core Features

Building Knowledge Graphs

Get started by building your first knowledge graph:
cd /path/to/your/repo
arc auth gh
arc build

Decision Trails

Discover the why behind every line of code:
arc why file path/to/file.py 42

Simulation

Predict the impact of changes before merging:
arc sim

LLM Integration

Connect knowledge graphs to AI assistants:
arc serve start

​
Troubleshooting & Support

Debug Mode

Enable debug logging with the --debug flag:
arc build --debug

Get Help