vantage.ai
See and control what your coding agent does.
65 followers
See and control what your coding agent does.
65 followers
Vantage is a free, open-source cost and usage monitor, guardrail and session log for coding agents.
Free
Launch Team
Framer AI AgentsDesign and publish professional sites with AI
Promoted
is this more of a review after the fact, or can I actually block a change before it happens? the "control" part is what I'm curious about.
being open source is the right call for something that sits between an agent and your keys, nobody should have to trust a closed binary for that. question on the guardrail side: does it block the action before the agent reads the secret, or does it let the read happen and just flag it in the session log after the fact? for us the scary case isn't the agent leaking a key in its output, it's the agent quietly using a live key to hit a paid API in a loop, so catching it before the call matters more than logging it afterward.
The .env secret detection is the thing that got me, that's such a real footgun with coding agents and nobody was catching it. Congrats on the launch, this fills a gap that's been quietly costing people money and security for a while now.