act101

act101 is a groundbreaking developer tool that fundamentally changes how AI coding agents interact with source code. It is the first dev tool that enables an AI agent to actually perform language-aware code work, such as refactoring Python, porting C to Rust, or migrating Ruby to Elixir. Instead of treating code as plain text and rewriting entire files, act101 gives agents access to a suite of precise, Abstract Syntax Tree (AST) based operations. This tool is designed for developers using AI coding assistants like Claude Code, Cursor, Codex, and OpenCode. Its core value proposition is that it allows AI agents to make surgical, intelligent changes to codebases with full awareness of the programming language's structure and semantics. This results in changes that preserve comments, formatting, and cross-file consistency, all while offering instant undo capabilities. act101 is built as a single, native Rust binary that runs a Model Context Protocol (MCP) server. It requires no indexing, caching, or cold start, meaning results are always fresh and your code never leaves your machine. It supports 163 programming language grammars, 183 AST refactor operations, 30 codebase analyzers, 15 query operations, 8 porting operations, and 10 pre-built agent skills. The tool is free for personal use and prioritizes security by having no plugin runtime, no package graph, and no supply-chain attack surface.

AST-Aware Refactoring Operations

act101 provides AI agents with 183 typed, AST-aware refactor operations. Unlike traditional file-based edits that rewrite entire files, these operations understand the structure of the code. The agent can call specific commands like extract-function, rename, move-symbol, inline, convert-to-dataclass, or organize-imports. These operations work across the entire codebase, maintaining cross-file consistency. Every operation creates an automatic checkpoint, and the agent can instantly undo any change that looks wrong. This precision, combined with support for 163 grammars, allows for safe and effective code transformation at a scale previously impossible for AI agents.

Cross-Language Porting State Machine

act101 introduces a state machine for end-to-end language migration. The porting process is driven by four key operations: port_contract to anchor the migration, port_inventory to enumerate every symbol that must move, port_order to resolve dependency ordering, and a port_manifest state machine with init, add, update, remove, and note steps. This structured approach allows an AI agent to port code between any two of the 163 supported grammars, such as C to Rust or COBOL to Java. It ensures that the migration is systematic, traceable, and complete, moving beyond simple text translation to a true structural transformation.

Comprehensive Codebase Analysis

The tool equips AI agents with 30 built-in codebase analyzers. These analyzers can assess cohesion, coupling, cycles, chokepoints, hotspots, dead code, layers, seams, clusters, surface area, fan balance, migration readiness, and type completeness. Before the agent makes a single change, it can get a complete structural map of the repository. This analysis allows the agent to make informed decisions, identify the most impactful areas for refactoring, and understand the dependencies and health of the codebase. It transforms the agent from a simple code editor into a sophisticated code architect.

Native MCP Server and Pre-Built Agent Skills

act101 runs as a native Model Context Protocol (MCP) server built into a single Rust binary. This design ensures it works seamlessly with any MCP-aware client, including Claude Code, Cursor, Windsurf, and Codex. There is no need for a plugin runtime or complex setup. On top of the raw operations, act101 provides 10 pre-built agent skills that compose these operations into common engineering workflows. These skills include architecture-audit, code-review, refactoring, code-navigation, code-generation, migration-assessment, boundary-analysis, change-impact, health-check, and architectural-refactoring. Users can invoke a skill with a simple command like /skill-name to execute a complex, multi-step task.

Large-Scale Code Refactoring

A development team needs to rename a core data structure across a multi-language monorepo. With act101, an AI agent can use the rename operation, which is AST-aware and understands the scope of the symbol. The agent renames the structure in every file, updates all references, and handles cross-file imports automatically. The team can review the changes with confidence, knowing that the tool preserved formatting and comments, and they can instantly undo any part of the refactoring if needed. This reduces a task that could take days to a matter of minutes.

Cross-Language Code Migration

A company has a legacy C library that needs to be ported to Rust for better memory safety and performance. An AI agent using act101 can initiate the porting process by first creating a migration contract. It then inventories every symbol in the C codebase, resolves the dependency order, and begins translating code through the port_manifest state machine. The agent handles the structural differences between the languages, ensuring the resulting Rust code is idiomatic and correct. This systematic approach makes complex migrations manageable and reduces the risk of introducing bugs.

Codebase Health Audits

An engineering lead wants to understand the architectural health of a growing Python and TypeScript project. They instruct an AI agent to run a health-check skill from act101. The agent uses the 30 codebase analyzers to assess coupling, cycles, dead code, and hotspots. It generates a report that identifies the most tightly coupled modules, the parts of the code with the highest change frequency, and potential architectural violations. This analysis provides the team with a clear, data-driven roadmap for where to focus their refactoring efforts.

Automated Code Review Assistance

A developer submits a pull request for a new feature. A reviewer uses an AI agent with act101 to perform a code review. The agent runs the code-review skill, which analyzes the changed code for structural issues. It can check for violations of layering rules, identify potential circular dependencies introduced by the new code, and assess the completeness of type hints. The agent provides a concise summary of its findings, highlighting areas that need attention without getting bogged down in stylistic nitpicks. This accelerates the review process and catches deep structural issues early.

How does act101 differ from other AI coding tools?

act101 is fundamentally different because it operates on the Abstract Syntax Tree (AST) of a codebase rather than treating code as plain text. Other tools typically rewrite entire files, which can break formatting, remove comments, and lack undo capabilities. act101 provides the AI agent with typed, surgical operations like extract-function or rename that understand the language's structure. This leads to more precise, safer, and more reliable code changes. It also supports a massive range of 163 languages and offers a state machine for cross-language porting, capabilities that no other single tool provides.

Is my code safe and private when using act101?

Yes. act101 is designed with privacy and security as a priority. It runs as a single native Rust binary on your own machine. There is no indexing, caching, or cold start, meaning your code is never stored or sent to external servers. The tool has no plugin runtime, no package graph, and no supply-chain attack surface. Your code stays on your machine at all times. The only network activity is for license verification, which may contact act101 servers. No telemetry is collected.

What AI coding assistants are compatible with act101?

act101 is built as a Model Context Protocol (MCP) server. This makes it compatible with any MCP-aware client. It works directly with Claude Code, Cursor, Windsurf, Codex, and OpenCode. Users can integrate it into their existing workflow without needing to switch tools. The built-in MCP server handles all the communication, so the AI agent can simply call the refactor, port, analyze, and query operations as needed.

What does "free for personal use" mean?

act101 is free for individual developers using it for personal projects, learning, or non-commercial work. This includes hobbyists, students, and developers working on open-source code. For teams or commercial use within a company, a paid license is required. The pricing page on the act101 website provides details on the different commercial tiers and their features. The free tier provides access to all the core operations and skills, making it a powerful tool for individual developers.