act101

Product Introduction

1. Overview: act101 is a native Rust binary that functions as a Model Context Protocol (MCP) server, providing AI agents with language-aware, Abstract Syntax Tree (AST) operations for code manipulation across 163 programming languages.
2. Value: It enables AI coding assistants (like Claude Code, Cursor) to perform precise, safe, and reversible code transformations—such as refactoring and cross-language porting—directly on the user's machine without a cloud dependency, preserving code formatting and comments.

Main Features

1. AST-Aware Refactoring Engine: Provides 183 typed refactor operations including extract-function, rename, move-symbol, inline, convert-to-dataclass, and extract-trait. These operations are grammar-aware, ensuring changes are syntactically correct and consistent across files, with automatic checkpointing and instant undo.
2. Cross-Language Porting State Machine: Features 8 porting operations that drive end-to-end language migrations (e.g., C to Rust, Ruby to Elixir) through a structured contract/inventory/ordering/manifest workflow, moving beyond simple transpilation to handle API and paradigm shifts.
3. Codebase Intelligence Suite: Integrates 30 analyzers for architectural insights (coupling, cycles, dead code, hotspots, migration readiness) and 15 query operations (skeleton, references, callers, control_flow) to give AI agents deep contextual understanding of a codebase before making changes.

Problems Solved

1. Challenge: AI agents traditionally perform code changes via whole-file rewrites, which strip comments, break formatting, lack undo capabilities, and are prone to errors due to a lack of AST understanding.
2. Audience: Software engineers, platform teams, and developers using AI-powered IDEs (Cursor, Windsurf) or agents who need to automate complex code maintenance, legacy modernization, and large-scale refactoring tasks safely.
3. Scenario: A team needs to assess and migrate a legacy COBOL codebase to Java. An AI agent, powered by act101, can analyze the codebase for migration readiness, inventory modules, and then execute a structured porting operation, transforming code while preserving business logic integrity.

Unique Advantages

1. Vs Competitors: Unlike generic AI chat interfaces or simple linters, act101 provides a deterministic, operation-based API for code transformation. It eliminates the token-heavy context of entire files, reducing token usage by ~85% on average compared to file-based agent operations.
2. Innovation: Its architecture as a single, dependency-free Rust binary with a built-in MCP server eliminates plugin runtimes and supply-chain attack surfaces. It parses code on-demand with no indexing or caching, ensuring analysis is never stale, and all data remains local for security and privacy.

Frequently Asked Questions (FAQ)

1. How does act101 integrate with my AI coding assistant? act101 runs as a local Model Context Protocol (MCP) server. Once installed, you configure your MCP-compatible client (like Claude Code, Cursor, or Windsurf) to connect to it, instantly exposing all refactor and porting operations to the agent within your IDE.
2. Is my code sent to the cloud when using act101? No. act101 is designed for local-first operation. Your source code never leaves your machine. The tool performs all parsing, analysis, and transformation locally, with license verification being the only potential external call.
3. What programming languages are supported for refactoring and porting? act101 supports 163 programming language grammars for analysis and refactoring. Specific porting operations are available for high-value migrations such as C to Rust, Ruby to Elixir, and COBOL to Java, with the framework extensible for other language pairs.