Enterprise gateway for autonomous AI agents with isolation and audit logs.
Isolate messaging channels, vault credentials, and audit agent actions securely.
Wirken is an enterprise gateway that acts as a secure switchboard between team messaging channels and autonomous AI agents. It is designed for security teams needing to contain compromised agents, enforce least privilege, and maintain tamper-evident audit trails.
It receives inbound messages from channels like Telegram, Slack, or Matrix, wakes up an ephemeral agent, and routes actions through permission checks and Docker-based sandboxes. Every tool call and file read is signed and written to a hash-chained audit log before execution.
Suits security and IT infrastructure professionals deploying autonomous AI agents in production environments. It does not suit teams wanting unconstrained agent execution or those unwilling to run Docker and configure isolated channels.
Tokens and API keys stay in a local vault, supplied only at the edge so models never see them.
Every agent action is written to a typed, hash-chained log before it executes.
Shell commands run inside temporary containers with no network access by default.
Connects to frontier APIs, local Ollama instances, or TEEs with per-agent spend limits.
Free under its license, with the only cost being the hardware and LLM API usage required to run it yourself.
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
Without Docker, the agent simply cannot run shell commands, while all other functionalities remain operational.
Each agent is capped by a defined per-agent spend budget enforced by a budget gate.
Credentials reside in an encrypted vault managed via an age-file backend or OS keychain.
See Zekai first in Google