Thinking about AI agent security, secrets management, and the MCP ecosystem.
Janee’s Runner/Authority split lets AI agents execute API calls without secrets ever touching their environment. The part talking to the agent never has secrets. The part holding secrets never talks to the agent.
Read more →OpenSeed runs autonomous AI agents that need GitHub and other API access. Here’s how they use Janee to give each creature exactly the permissions it needs — with per-agent isolation, audit logging, and zero credential exposure.
Read more →AI agents are calling APIs on your behalf. But how do they get the keys? Most setups paste credentials directly into prompts or environment variables the agent can read. Here’s why that’s dangerous — and how a purpose-built secrets manager changes the game.
Read more →