MCP v0 is live: Connect Versuno directly to your coding agent

Public API v0 gave teams a programmable foundation. CLI v0 brought that power into local file workflows.

Now MCP v0 brings Versuno into the place where many developers work all day: their coding agent.

This release means your agent can discover, load, pull, and push Versuno assets without you leaving your editor.

Why this matters

Teams using AI tools often hit the same problem.

The best prompts, contexts, personas, and skills exist, but they are not always available at the exact moment the agent needs them.

So people fall back to repeating instructions in chat, copying content manually, and losing consistency across tasks.

MCP v0 fixes that by connecting your agent to your intelligent context layer.

Your agent can now:

• find the right asset when needed
• load that asset into the current conversation
• sync assets to local markdown files
• push improved versions back to Versuno

Why we built MCP v0 after API v0 and CLI v0

The order was intentional.

API v0 gave us a stable base. CLI v0 proved the local sync model for real workflows. MCP v0 builds on both, so agent actions map to the same trusted asset and versioning system.

This keeps everything aligned:

• one source of truth in Versuno
• one version history across UI, CLI, and MCP
• one workflow from agent idea to versioned update

What MCP v0 can do today

MCP v0 ships with five core tools:

• list_assets: browse assets by type with metadata only
• search_assets: full-text search across assets by relevance
• get_asset: load full asset content into agent context
• pull_asset: write an asset to .versuno as markdown with frontmatter
• push_asset: create a new version from a local .versuno markdown file

This gives you a practical loop in real development sessions:

1. Ask the agent to find relevant assets for the task.
2. Load the right skill or context into the session.
3. Pull to disk when you want to edit locally.
4. Push changes back as a new version.

Quick start

The fastest way to get started is the MCP docs:

https://docs.versuno.ai/mcp/overview

It includes setup for Cursor, Claude Code, Claude Desktop, VS Code, Windsurf, Zed, and other MCP-compatible clients.

Example setup

Add Versuno MCP to your MCP client config:

{
  "mcpServers": {
    "versuno": {
      "command": "npx",
      "args": ["-y", "versuno-mcp"],
      "env": {
        "VERSUNO_API_KEY": "uk_live_xxx"
      }
    }
  }
}

Then restart your client and ask:

"List my Versuno skill assets."

Your agent can call list_assets, show results, and continue from there.

Example agent workflow

A common flow looks like this:

• "Search Versuno for onboarding context"
• "Load the best match"
• "Pull it to disk so we can edit it"
• "Push it back with changelog: tighten onboarding steps"

Under the hood, that can map to:

• search_assets
• get_asset
• pull_asset
• push_asset

All connected to the same account and version history you already use.

Security and control

MCP v0 was built with clear boundaries.

• push_asset only accepts markdown files inside .versuno
• file size is capped to prevent unsafe payloads
• API keys are validated strictly
• you keep control through your MCP client's tool approval settings

In short, the agent gets useful access, but not unlimited access.

Who should use MCP v0 now

MCP v0 is a strong fit if you:

• use coding agents daily and want consistent context
• maintain shared prompts, skills, or personas across projects
• want faster iteration without losing version control
• want agent workflows connected to a real system of record

Start now

If you already use API v0 or CLI v0, MCP v0 is the natural next step.

Connect your client, run your first asset search, and let your agent work with the same context layer your team already trusts.