Markdown Editor

Research, write, plan, document in a local, agentic, WYSIWYG AI-powered markdown editor

Approve AI red/green changes.

Plan mode — v2 migration plan with goals, timeline, and mermaid diagram

Use AI and Claude Code and Codex to research, write, inspect your code, and edit markdown

Collaborate with AI and agents in your planning and spec documents, with your code, and in your workflow.

Editable WYSIWYG and native markdown

Editable WYSIWYG and native markdown

Work in a first-class fast markdown editor

Red / Green diffs show AI changes for your approval

Red / Green diffs show AI changes for approval

Review every change the AI makes. Accept, reject, or edit

Text, lists, tables, code

Text, lists, tables, code

Full markdown support across every content type

Images

Images

Embed images inline. AI can see images.

File history with restore

File history with restore

Every document version is saved. Browse history and restore any prior state of your file.

Drawing & Diagrams

Create and edit drawings and diagrams visually with your agents

Use Mermaid and Excalidraw to diagram systems, flows, and architecture

AI-generated Mermaid diagrams

AI-generated Mermaid diagrams

Describe a flowchart, sequence diagram, or class diagram in plain language and Claude Code and Codex generates the Mermaid code instantly.

Live rendered in markdown

Live rendered in markdown

Mermaid diagrams render inline in your markdown files with a live preview. Edit the Mermaid yourself or with your agent.

Draw with Excalidraw

Draw with Excalidraw

Sketch architecture diagrams, user flows, and system maps on an infinite canvas with Excalidraw edit by hand or with AI.

AI-powered drawing

AI-powered drawing

Describe what you want and Claude Code and Codex places, connects, and styles Excalidraw elements for you. Iterate conversationally or with the drawing tools.

Agent Sessions

Manage your agents' work across parallel sessions in a UI

Run multiple agents simultaneously. Organize them. Find, resume, branch. Connect sessions to files and files to sessions.

Run multiple sessions in parallel

Run multiple sessions in parallel. Alert when done.

Kick off agents across different tasks and get notified as each one completes.

Tie sessions to files and files to sessions

Tie sessions to files and files to sessions

Every session is linked to the files it read and wrote. Optionally open edited files and review.

Search and resume your persistent AI sessions

Search & resume your persistent AI sessions

Sessions are saved and searchable. Pick up any conversation exactly where you left off. Branch. Archive.

Developer Features

Code with your agents, iterate, review changes, manage files, use AI to commit

You are coding in a new way. You need a purpose-built environment for agentic coding. Chat, read, edit, and review — all without leaving the workspace.

Instruct your coding agent

Instruct your coding agent.

Talk to Claude Code and Codex in a full chat interface. Describe what to build, @ mention files, use / commands and skills, and watch it work.

Read and human edit code in Monaco editor

Read and human edit code in Monaco editor. Use AI to edit.

Full Monaco editor with syntax highlighting. Edit manually or ask AI to make targeted changes.

See files changed by agent session

See files changed by agent session

A sidebar lists every file the agent read or wrote. Review what changed before you commit.

Integrated terminal

Integrated terminal

Run commands, scripts, and build tools without leaving Nimbalyst. Full terminal access alongside your editor and agent sessions.

Git management

Git management

View diffs, stage changes, and manage branches from a visual git interface built into your workspace.

AI-assisted git commits

AI-assisted git commits

Claude Code and Codex reviews your staged changes and drafts a commit message. Review, edit, and commit in one step.

Worktree support

Worktree support

Work on multiple branches simultaneously with git worktrees. Each worktree gets its own editor context and agent sessions.

Planning & Tracking

Track your document status and ideas, plans, and to-dos in markdown

WYSIWYG and stored as markdown frontmatter so humans and any agent can read and update while working.

Workflow & status in your plan

Workflow & status in your plan

Kanban board views, priority, progress, and owner — all stored in your plan document as markdown frontmatter.

Cross-plan status in markdown

Cross-plan status in markdown

See status across all your plans in one view. Decisions, bugs, tasks, and ideas tracked in one dashboard.

To-dos, tasks, ideas in your doc

To-dos, tasks, ideas in your doc

Tag items inline as @task, @idea, @bug, or @decision. They surface in your sidebar and plans automatically.

Agent Integration

Multi-agent workspace for Claude Code, Codex, and more

Run Claude Code and Codex side by side. Use the best agent for each task, with shared context, MCP servers, and a unified workflow.

Claude Code

Claude Code

Built around Claude Code. Chat, plan, code, and review

Codex

Codex

Run OpenAI Codex alongside Claude Code. Use the best agent for each task.

/ commands and skills

/ commands and skills

Extend your agent with custom slash commands and skills. Project-specific or personal, versioned with your repo, and shareable with your team.

MCP support

MCP support

Connect your agents to external tools and services via Model Context Protocol. Linear, GitHub, Slack, databases, and more — all accessible from your sessions.

CLAUDE.md / Agents.md

CLAUDE.md / Agents.md

Configure agent behavior with project-level instruction files. Set context, constraints, and conventions that your agents follow automatically.

Open & Secure

Its your content and your data.

Everything is open, local, and yours.

Open storage of content and status in markdown

Open storage of content and status in markdown

All content and metadata lives in plain markdown and other standard files on your filesystem. No proprietary formats, no lock-in.

Use your own AI APIs

Use your own AI APIs

Bring your own API keys for Claude, OpenAI, and more. Run entirely locally with LM Studio. You're always in control.

We don't collect your data

We don't collect your data

Your data lives locally in your file system and is synced via GitHub if you choose. We collect usage data only, and you can opt out.

Permission System

Permission System

Fine-grained control over what agents can do. Set trust levels, permission modes, allowed tools, and filesystem access per project.

SOC 2 Certified
Nimbalyst SECURE

Nimbalyst is SOC Certified

Nimbalyst is SOC 2 Type 2 Certified

Nimbalyst is SOC 2 Certified

Nimbalyst is SOC 2 Type 2 certified. Enterprise-ready security you can trust.

Nimbalyst: Agent-native visual workspace for Builders