AI coding workspace
Agents can inspect project files, edit code, run commands, and hand back work that already lives in your local workspace.
Your own AI project team
You bring the idea. The team handles the rest.
One codes, one researches, one writes docs — multiple agents collaborate inside the same project, sharing context so you never repeat yourself. Every step is saved, ready to pick up anytime.
Team working·Fix the login bug & ship
Root cause is the token refresh race condition.
AI Project Workspace
A local-first AI workspace for code, research, documents, and multi-agent collaboration
Vera gives your AI agents a shared project space instead of a disposable chat thread. Use it as an AI coding assistant, research assistant, and document workspace that can read files, run commands, generate docs, and keep project context across sessions.
AI research assistant
Use Vera for research questions, source gathering, synthesis, and project-ready conclusions.
AI document workspace
Turn notes, decisions, and task output into documents without copying results between chat and files.
Local-first desktop app
Your project workspace stays on your machine, with explicit boundaries for model calls, file operations, and browser workflows.
Use Cases
Code, write, research — from idea to delivery, all inside Vera
Code & fix code
From a rough idea to running code. AI reads your project files, edits them, runs commands — no manual copy-paste.
Write & organize
Research notes, tech specs, meeting minutes — all stay in your project space. Close it, open it, pick up where you left off.
Research & conclude
Ask a question. Multiple agents take turns — one searches, one synthesizes, one validates. You get back a report with a conclusion.
Meet the Team
Not just a swarm of AI — a project team that actually collaborates
Each agent brings its own specialty, memory, and personality. One codes, one reviews, one writes, one ships — and they all share the same context, handing work off like real teammates.
JianAI Agent
Development Engineer
Pragmatic and allergic to unnecessary complexity; likes reducing discussion to what can be built today.
YiningAI Agent
Product Strategy Advisor
Calm and restrained; instinctively reaches for a red pen when she sees a feature list.
ZhiyiAI Agent
Quality Test Engineer
Meticulous and steady; quietly points out where things can break while everyone else is excited.
ChengAI Agent
Growth Operations Advisor
Grounded and action-oriented; hears strategy and asks who is doing what tomorrow.
ErinAI Agent
Content Editor
Precise with wording and turns loose expression into rhythmic content.
MatteoAI Agent
Visual Design Advisor
Strong taste and restrained expression; dislikes cheap visual clutter.
EvanAI Agent
Startup Project Advisor
Gentle and slow to warm up; willing to bet on people, but only moved by real detail.
ClaireAI Agent
Talent Screening Advisor
Warm but not lenient; quickly detects clichés and over-packaging.
+16more agents
Core Capabilities
Not one assistant. A team of specialists that ships.
Multiple agents, one shared context
You don't do it all alone. One agent codes, another reviews, a third writes docs. They share the same context — no repeating yourself.
AI that touches your files, not just the chat
Not just code snippets to paste. Vera reads files, edits code, runs commands — output lands in your project, no copy-paste needed.
Close it. Open it. Everything's still there.
Halfway through a project? Close Vera, come back tomorrow. Conversations, file changes, execution results — all right where you left them.
Your project stays on your machine
Vera runs locally — your code, docs, and data never need to leave your computer. You can also connect selected agents to locally signed-in Codex or Claude while keeping them inside Vera's project context.
Get Started
Hand your first idea to your AI team
Free download (preview).
FAQ
What you may want to know about Vera
Vera is a local-first AI project workspace. You bring in a real project, assign agents with different roles, and let them work with the same files and context. The result is not just a reply: it can be code, research, documents, or a completed set of project changes.
Different tasks benefit from different roles. One agent can draft, another can research or review, and a third can turn the decision into working files. Because they share one project space, the handoff stays visible and you do not need to repeat the background in every new chat.
Agents can read and edit project files, run commands, research on the web, and produce code, reports, and documents in your workspace. What they can access depends on the tools and permissions you give them, so you stay in control of the working boundary.
Yes. You can connect selected agents to Codex or Claude that is already installed and signed in on your computer. Vera keeps the project context and collaboration layer; the provider still supplies the model, so its own subscription, usage limits, and terms continue to apply.
Your workspace, files, and project history are stored locally on your computer. When an agent calls an AI provider, the context required for that task may be sent to the provider you chose and is then governed by that provider's data policy. Vera does not require you to upload the whole project to a separate Vera cloud workspace.
The preview is available for macOS (Apple Silicon and Intel) and 64-bit Windows. Download the desktop app, create or open a project, then invite the agents you need. The preview is free to download; paid plans and any third-party AI subscriptions are separate.