AgentSecrets
Zero-knowledge credential infrastructure built for AI agents and teams
AgentSecrets is a credential management platform designed specifically for AI agents, preventing prompt-injection-based credential theft by ensuring agents never directly access API key values. It uses a local HTTP proxy to inject credentials at the transport layer, combined with zero-knowledge encrypted cloud sync, team workspaces, MCP integration, audit logging, and agent identity management. Unlike traditional secret managers where agents retrieve and hold credentials in memory, AgentSecrets keeps credentials structurally unavailable to the agent layer.
A local HTTP proxy intercepts agent API calls, resolves credentials from the OS keychain, injects them at the transport layer, and returns only the API response — so the agent never sees the actual credential value.
developers and teams building or using AI agents that make authenticated API calls
Background.
- Status
- launched
- Business model
- freemium
- Company
- The Seventeen
Similar projects.
Editorial take on the space this project sits in — momentum signals, adjacent moves, our call on whether the wedge is real. Get pinged when we publish a new read or when the landscape shifts.
Have a take on this space?
Tell us what you’d build differently, where you think the incumbents miss, or what we’ve gotten wrong about this project. Comments + reactions are coming soon.