AgentAvow Trust MCP server

AgentAvow is a safety check for the tools and add-ons you connect to your AI. You point it at a GitHub project, an MCP server, or an npm or PyPI package, and it gives back a letter grade from A+ to F. It is handy if you keep hearing about AI tools and want a quick, plain answer to "is this one okay to use?"

What this MCP server does

You ask your AI something like "is the chalk package safe?" Your AI sends that question to the AgentAvow helper. The helper looks up the project and runs a safety scan on it. It checks things like whether secrets are left in the code, how the code behaves, and what it depends on. Then it hands back a grade, a score, and a short list of what it found, and your AI tells you the result in the chat.

Click to zoom

What you can do with it

• Get a letter grade for an MCP server before you connect it
• Check an npm or PyPI package before you install it
• Scan a GitHub project and see what was flagged
• See the exact line or file each finding points to
• Compare the safety scores of two tools side by side
• Look up a wallet address and scan the project it links to
• Ask for a fresh scan when you want the latest result

Try asking your AI

• “Scan the npm package chalk and tell me its safety grade”
• “Is this MCP server safe to connect? Here is its GitHub link”
• “What did the last scan of that repo find?”
• “Give me a fresh scan of this package, not the cached one”

What it gives back to you

You get a letter grade from A+ down to F, plus a number score out of 100. You also get a short list of findings, each with a severity level and a pointer to the exact line or file. Your AI sums this up in the chat, so you can decide whether to go ahead. Every result also comes with a signed proof you can check later if you want.

Before you start

Who it's for

Anyone who connects AI tools and add-ons at work and wants a quick safety check before saying yes.