act101

act101 is a treasure map for developers who want their AI agents to stop just talking about code and start actually wielding the chisel. It is the first dev tool that lets an AI agent perform language-aware refactoring and porting across an astonishing 163 programming languages. Think of it as a master key that unlocks a hidden vault of 183 Abstract Syntax Tree (AST) refactor operations, 30 codebase analyzers, 15 query operations, and 8 porting operations. Instead of your AI agent blindly rewriting entire files and losing comments or breaking formatting, act101 gives it a set of precise, surgical tools. It is a Model Context Protocol (MCP) server built as a single native Rust binary. It works seamlessly with Claude Code, Cursor, Codex, and OpenCode. The tool is designed for engineers who are tired of manual, error-prone code migrations and who want to automate complex refactoring with the confidence of a seasoned explorer. Your code never leaves your machine, and there is no indexing, caching, or cold start. It is a direct line to deep code intelligence, making every AI-driven change a deliberate, undoable, and consistent step forward.

163 Grammars in One Binary

This is the Rosetta Stone of programming languages, all packed into a single Rust binary. act101 understands 163 different grammars, from the ancient scripts of COBOL to the modern dialects of Rust, Python, TypeScript, and Go. This means your AI agent can navigate, analyze, and transform code across an entire kingdom of languages without needing separate plugins or runtimes. It parses on call, so results are never stale, and there is no heavy indexing to slow down your quest.

183 AST Refactor Operations for AI Agents

This is the treasure chest of surgical tools. act101 provides 183 distinct operations that an AI agent can call to reshape code at the abstract syntax tree level. Think of operations like extract-function, rename, move-symbol, inline, convert-to-dataclass, extract-trait, and organize-imports. Every operation is typed and AST-aware, meaning the agent understands the structure of the code, not just its text. Each change is automatically checkpointed, giving you the power of instant undo if the path leads to a dead end.

8 Porting Operations for Cross-Language Migration

This is the enchanted bridge between languages. act101 introduces a state machine for porting code that uses four key operations: port_contract, port_inventory, port_order, and a port_manifest state machine. The agent first anchors the migration with a contract, then inventories every symbol that must move, resolves the dependency ordering, and tracks progress step-by-step. This turns the chaotic art of porting C to Rust or Ruby to Elixir into a methodical, guided expedition.

30 Codebase Analyzers and 10 Pre-Built Agent Skills

This is the map and the compass. act101 includes 30 analyzers that probe the hidden depths of your codebase, measuring coupling, cycles, dead code, hotspots, layers, and migration readiness. It also comes with 10 pre-built agent skills like architecture-audit, code-review, refactoring, and migration-assessment. These skills compose the raw operations into powerful workflows, so you can simply invoke a skill and let the agent begin its automated treasure hunt.

Automating Complex Refactoring in a Monorepo

In a sprawling monorepo with thousands of files across Python, TypeScript, and Go, a developer can use act101 to have an AI agent perform a massive rename of a core library symbol. The agent uses the rename operation to update every reference across all files, ensuring cross-file consistency. The AST-aware nature of the operation means it correctly handles scoping and avoids renaming local variables. The automatic checkpointing allows the developer to review and undo the change instantly if the ripple effect is too large.

Porting Legacy COBOL to Modern Java

A financial institution needs to modernize a critical COBOL transaction processing system to Java. With act101, an AI agent begins by running a migration readiness analyzer to map the codebase. It then uses the port_contract operation to define the migration scope, followed by port_inventory to list every COBOL data structure and procedure that must be ported. Using the port_order operation, the agent resolves the dependency graph and begins generating Java classes, methods, and logic, tracking each step in the port_manifest state machine.

Analyzing Codebase Health for a Security Audit

A security team is tasked with auditing a large Rust codebase for architectural weaknesses. They invoke the architecture-audit skill in act101. The agent runs the coupling and cycles analyzers to identify tightly coupled modules and circular dependencies. It then uses the hotspots analyzer to find files with high churn and complexity. The agent produces a structured report with a graph of the module boundaries, pinpointing the most dangerous areas for the team to investigate.

Automated Code Review with Structural Insight

During a pull request review, a developer asks an AI agent to perform a code review using act101. The agent uses the code-review skill, which first runs the dead code analyzer to find unused functions in the new code. It then uses the change-impact analyzer to see which existing modules might be affected by the changes. The agent then uses the skeleton and references query operations to understand the new code's structure and dependencies, providing a review that goes far beyond surface-level syntax checks.

How does act101 keep my code safe and private?

act101 is built as a native Rust binary that runs entirely on your machine. It does not send your code to any external servers. There is no telemetry, and the only network contact is for license verification. All parsing, analysis, and transformations happen locally. Your code stays on your machine, making it a secure choice for even the most sensitive codebases.

What AI coding agents are compatible with act101?

act101 is built on the Model Context Protocol (MCP), making it compatible with any MCP-aware client. This includes Claude Code, Cursor, Windsurf, Codex, and OpenCode. The tool exposes its operations through a built-in MCP server, so any agent that speaks the MCP language can immediately access the 183 refactor operations, 30 analyzers, and 8 porting operations.

What does it mean that act101 uses AST refactor operations instead of text-based operations?

Traditional AI code tools work on raw text, meaning they rewrite entire files to make changes. This often destroys formatting, removes comments, and has no concept of code structure. AST refactor operations work on the Abstract Syntax Tree, which is the structural representation of the code. This allows the agent to perform precise, surgical changes like extracting a function or renaming a symbol without touching anything else. It also enables cross-file consistency and instant undo.

Is act101 free for personal use, and how does licensing work?

Yes, act101 is free for personal use. For commercial or team use, you can check the pricing page on the website. The tool may contact our servers for license verification, but this is minimal and does not involve any code being sent. The single binary approach means there is no complex installation or dependency management.