Agent runs with visible controls
Run Codex or Claude Code inside a signed macOS app while keeping authentication, permissions, model choice, and session state visible during the run.
memi
memi is a signed macOS workbench for product designers to run Codex or Claude Code with project memory, design-system context, receipts, and Figma/FigJam handoff. It is built to keep agent runs inspectable and reusable inside a project-first desktop workflow.
Product overview
memi is a macOS workbench for product designers who want to run coding agents with design-system context attached. The homepage describes it as a signed app where Codex and Claude Code run alongside project memory, receipts, and Figma/FigJam handoff.
The product centers on keeping design work inspectable. Rather than hiding decisions in prompts, memi lets teams turn research, tokens, artifacts, and notes into editable memory, then export planning source and design handoff material that can be reviewed before external sync.
Core capabilities
Editable project memory
Store decisions, tokens, research, notes, and reviews as editable project memory instead of leaving them buried in prompts or chat history.
Run receipts and log access
Review prompts, plans, tools, files, and cost as receipts, with a compact run spine and access to the raw log when audit detail is needed.
Board-ready planning exports
Export local Mermaid and FigJam-ready planning source so board work can stay inspectable before any external sync is approved.
Figma bridge for design context
Pull design-system source from Figma when needed, including tokens, components, trees, and screenshots for contextual runs.
Design-system memory layer
Use the bundled design system reference to work from tokens, primitives, patterns, and component mappings that are intended for autonomous reskinning.
Practical use cases
Iterate on a design system
Open a product design project, run an agent with the current design system attached, and keep the result in project memory so the next pass can reuse decisions and tokens.
Audit an agent-assisted change
Run Codex or Claude Code while reviewing prompts, plans, tools, files, and cost as receipts, which makes it easier to audit how a result was produced.
Add design context to a coding task
Pull Figma tokens, components, trees, or screenshots into the workflow when a run needs current design source rather than only prompt text.
Create inspectable planning artifacts
Prepare board-ready planning material in Mermaid or FigJam-friendly source, review it locally, and only then approve any external sync.
Maintain shared project memory
Capture research, specs, notes, and decisions as editable memory so a team can build a readable project record instead of relying on scattered chat threads.
Pros and Cons
Pros
- Combines agent execution, project memory, receipts, and design handoff in one desktop app.
- Keeps prompts, tools, files, and cost inspectable for review and auditing.
- Uses editable Markdown/YAML memory instead of hidden prompt state.
- Includes a Figma bridge and design-system reference for context-aware runs.
- Provides local planning exports before external sync, which supports review-first workflows.
Cons
- The published pages do not show pricing details, plan tiers, or a working pricing page.
- The source does not describe support for non-macOS platforms.
- Some workflows, such as external sync and Figma use, appear approval-gated, so not every output is automatically pushed to downstream tools.
FAQ
What is memi built for?
memi is a signed macOS app, and the homepage describes it as a workbench where you open a project, choose an agent, and run it with project memory and design-system context.
Which agents does memi run?
The source describes Codex and Claude Code as the primary agents. The app keeps auth, permissions, models, and session controls visible while the run is in progress.
What do the receipts and run spine show?
The homepage says memi can inspect prompts, plans, tools, files, and cost as receipts, and that the run spine stays compact until you need the raw log.
What kind of output can I keep from a run?
The product mentions design memory in Markdown and YAML, plus exports for local Mermaid and FigJam-ready planning source before external sync is approved.
Does memi only run on macOS?
The site presents memi as a macOS Studio app, and the docs and changelog reference a Figma bridge and a local bridge port for design-source connection. The available pages do not describe a Windows or web app equivalent.
Quick Facts
- Category
- AI workbench for product designers
- Platform
- macOS
- Primary agents
- Codex and Claude Code
- Source domain
- memoire.cv
- Product model
- Signed desktop app with project memory and handoff tooling
- Open source
- Yes, as stated on the homepage
Alternativas a memi
Redline
Redline is a budgeting tool for Claude Code that paces sessions to stay within time, token, cost, or plan-percentage limits. It uses Claude Code’s native hooks and statusline to help sessions finish with a usable result instead of stopping abruptly.
Macuse
Macuse is a macOS app that lets AI assistants control native Mac apps through MCP and use Computer Use for any app. It works with clients like Claude Desktop, Cursor, and Raycast, and keeps automation on-device.
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Goldfish
Goldfish is an AI memory app for macOS that helps you reply, write, summarize, and continue using the context already on your Mac. It stores memory locally and writes in your tone.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
Biji
Biji es una plataforma versátil diseñada para mejorar la productividad a través de herramientas y características innovadoras.