Cadenya
Freemium
Developers building agentic workflows
Mind-expander
FreeOpen SourceSelf-HostedAgentic
AI coding agents are only as useful as their context — and when a codebase spans dozens of modules, the agent is flying blind through a flat file tree. mind-expander turns that file tree into a source-backed graph on an infinite canvas, giving both you and the agent a shared map of modules, types, calls, and dependencies.
Similar tools
CadenyaFreemiumCadenya is an agent runtime built around a layered model: tools connect once through MCP…
WorldsFreemiumWorlds spins up a sealed twin of the external systems your agent touches — Stripe and Zendesk…
Brain MemoryFreeBrain Memory stores agent decisions as Markdown files with YAML frontmatter, organized in a…
JevProxyFreemiumThe proxy sits in front of your Cursor or Claude Code agent fleet and classifies incoming tool…
Get This Tool
License: Apache-2.0 Any use incl. commercial
Local-run terms: Apache-2.0 allows free use, modification, and distribution for commercial and private purposes, with attribution required and liability/warranty disclaimers.
Pricing
- Model
- Free
Summary
The agent drives the canvas: it can run `npx mind-expander` in the background, load skill integrations, and build guided tours through architecture. You see the same graph the agent is reasoning about, which means review decisions and refactor plans are grounded in actual dependency structure — not the agent's approximation of it. That shared view is the differentiator. The ceiling arrives with language support: Rust and TypeScript are covered, the docs describe more language frontends as planned. Teams whose core services are in Go, Python, or Java will hit that wall on day one.
Bottom line: Pick this when your AI-assisted refactor lives in a Rust or TypeScript codebase and you need both you and the agent working from the same dependency graph — but plan a different approach when your stack is outside those two languages.
Find your fit
Before you choose Mind-expander
More options in Agent Frameworks. Compare pricing, platforms, and features side by side.
Brain Memory
Free · Open source
Long-term continuity in multi-session coding projects
Pros
- Source-backed dependency graph generated from actual code rather than agent inference, so the agent's architecture reasoning is grounded in real module relationships instead of reconstructed approximations that break on unfamiliar patterns.
- Agent-steerable canvas with guided tour support, which means an AI agent can walk a developer through an unfamiliar codebase interactively — replacing a static wiki page that goes stale the week after it's written.
- PR and commit impact visualization scoped to the actual nodes changed, so reviewers see cross-boundary effects in the dependency graph without manually tracing every import chain.
- Fully open-source under Apache-2.0 with no paid tier, so the tool can be self-hosted and extended without a licensing negotiation when your team needs a custom language frontend or a different rendering surface.
- First-class Claude agent integration via a dedicated skill directory and hooks, which means agent setup follows a documented protocol rather than a trial-and-error prompt engineering session.
Cons
- Language support is limited to Rust and TypeScript at the time of publication — teams with Go, Python, Java, or mixed-language services cannot use the graph features at all. There is no workaround short of contributing a new language frontend. Teams in those stacks will evaluate a different static analysis or diagramming tool from day one.
- No API surface is exposed, so integrating mind-expander into a CI pipeline or a custom agent harness outside the supported skill integration requires forking the project and building that surface yourself — at which point you are maintaining a fork.
- The agent integration is documented specifically for Claude; teams running GPT-4, Gemini, or a self-hosted model will find the skill directory and hooks are Claude-shaped, and adapting them to a different agent framework is undocumented and likely manual.
About
- Platforms
- Web (browser-based), CLI (npx)
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- June 9, 2026
Best For
Who it's for
- Development teams using AI coding agents for architecture work
- Code reviewers analyzing multi-file changes and architectural impacts
- Developers planning large refactors with AI assistance
- Teams handling Rust and TypeScript codebases
What it does well
- AI agent-guided code architecture review and planning
- Understanding codebase structure and module dependencies before refactoring
- Visualizing pull request changes across system boundaries
- Onboarding developers by walking through codebase architecture interactively
- Tracing impact of commits on system structure and dependencies
Integrations
Compare Mind-expander
Frequently Asked Questions
- Is Mind-expander free?
- Mind-expander offers a free option. Hosting, model usage, or optional services may have separate costs; check the pricing details and official terms.
- Is Mind-expander open source?
- Yes. Mind-expander is open source.
- Can I self-host Mind-expander?
- Yes. Mind-expander supports self-hosting on your own infrastructure.
- What platforms does Mind-expander support?
- Mind-expander is available on: Web (browser-based), CLI (npx).
People also compared
Curated lists that include this category
AI agents planning a refactor can describe a module graph accurately in simple cases and confidently wrong in complex ones — because they reconstruct structure from text, not from the actual source. mind-expander closes that gap. It parses a codebase into an infinite-canvas graph: modules, types, call relationships, dependency edges, and the specific nodes touched by a commit or PR. Agents connect through a skill integration, invoke the tool in the background, and can steer the visualization directly. You and the agent share the same source-backed workspace, not separate mental models.
The differentiating capability is agent-directed guided tours. An agent can walk through codebase architecture interactively — highlighting the path from an entry point through its dependency chain, or isolating the blast radius of a proposed change. This is not a static diagram the agent describes; the vendor describes agents actively creating and steering these tours, which means onboarding a developer or scoping a PR review becomes a collaborative session with the agent rather than a briefing followed by manual verification.
The tool fits best on teams running AI-assisted architecture work where the codebase is Rust or TypeScript — pre-refactor scoping, PR review across system boundaries, and developer onboarding where a new team member needs to understand module ownership fast. It breaks immediately for teams whose primary services are in Go, Python, Java, or any language outside the two supported frontends. The project is Apache-2.0 licensed, fully open-source, self-hosted via a local server, and carries no paid tier — so the ceiling is entirely about capability, not cost.
Distribution is through npm (`npx mind-expander`) alongside a Rust core built with Cargo. The repository includes a `.claude/` directory with hooks and a dedicated `skill/` directory, indicating Claude-specific agent integration is a first-class workflow rather than an afterthought. An `AGENTS.md` file documents agent interaction protocols directly in the repo.
Contribute
- Work on Mind-expander? Claim this listing.
- Embed a badge for your README or site.
Sign in to rate this tool or suggest an edit.