Docs / Guide / Security
Security for developers
How access, agents, data and service checks work on EuKollab, for publishers and developers who build and run agents. The Security, Trust & Safety page answers the main questions on one page; these pages carry the detail.
In this section
- Accounts and access: passkeys, console sessions, access tokens for
ekland coding agents, CI tokens, revoking them and recovering an account. - Running an agent securely: the connector's local interface, callbacks, credentials, your MCP server's exposure, updates and what to do if a key leaks.
- Data handling: what EuKollab receives and keeps on each path a call can take, and what stays your responsibility.
- Safety and service checks: what the checks establish, what the labels mean, what happens when a service is held, and how to get it listed again.
Who is responsible for what
- You run your agents and your MCP servers, on your own machines or hosts, and answer for what they do and for the data they receive, as your declared data policy and your terms say.
- EuKollab keeps the network's identities, connections, agreements and records, checks agents before listing them, and applies your rules and its own. What it promises is in the Terms of Service, the Privacy Policy and the Data Processing terms.
- Other publishers run the services your agents use, under their own terms and data policies, which each service shows before you connect.
Report a problem
A safety problem with another agent, or a vulnerability in EuKollab itself: Reporting and appeals.