Learn · Enforcement
Operational boundaries for AI agents, enforced at runtime
A config file that lists what an agent should do is a wish. An operational boundary enforced at runtime is a wall. This covers what an operational boundary is, and why where you enforce it matters more than where you declare it.
What an operational boundary is
An operational boundary is the precise envelope of what an agent may do: which operations it can perform, which resources it can touch, which systems and audiences it can reach, and under what conditions. It is the concrete expression of least privilege for one agent. A good boundary is specific, "read records of type X in system Y", rather than broad, "access the database." The narrower and more explicit it is, the smaller the damage any single agent can cause.
Config time vs runtime
There is a world of difference between declaring a boundary and enforcing one. A boundary written in a config file and read once at startup tells you the intended behavior, but nothing stops the agent from doing more if its credentials permit it. Runtime enforcement means the boundary is checked on every request, at a point outside the agent, right before the action would happen. Config-time declarations describe; runtime enforcement prevents. For agents driven by models that can be nudged off course, only prevention counts.
If the only thing holding an agent inside its boundary is the agent itself, you do not have a boundary, you have a hope.
How runtime enforcement works
Enforcement lives at Oathera Agent Enforcement in front of your tools. On each enforced request Oathera Agent Enforcement verifies the agent's identity and proof, checks the tool and its arguments against its configuration, and asks the Oathera Identity Service for a signed boundary decision. In-scope requests proceed; out of scope requests are refused and recorded. Because Oathera Agent Enforcement sits outside the agent and re-checks every time, every enforced call is checked again, including whether the agent and its principal are still active. Two limits: with the default-deny setting off, which is the default outside production, a tool the server lists but you haven't configured passes through unenforced and is logged; and calls that never reach Oathera Agent Enforcement aren't seen. Combine this with filesystem and network confinement from the sandboxing tools you already run, so the boundary holds even for actions that never reach Oathera Agent Enforcement.
Oathera Agent Enforcement refuses unknown tools, tools you haven't configured when its default-deny setting is on, and arguments outside a tool's allow-list, and the boundary decision checks each agent's list of permitted services.
How to set it up
- Derive the boundary from the task. List only the operations, resources, and audiences the agent needs.
- Express it as policy. Keep it in version control, reviewed like code; see Oathera's own rules.
- Enforce at Oathera Agent Enforcement per request. Verify identity, evaluate the boundary, deny by default.
- Confine the runtime. Use your sandboxing tools so filesystem and network access match the boundary too.
- Record and expire. Capture boundary decisions and rely on a short identity-token lifetime so authority never outlives the task.
Oathera supplies the identity, Oathera Agent Enforcement enforcement point and the boundary decision. Denying by default is an Oathera Agent Enforcement setting (GATEWAY_DEFAULT_DENY), off by default outside production, and the boundary decision checks the agent's list of permitted services. See features and the use case.
FAQ
What is an operational boundary for an AI agent and how do you enforce one at runtime rather than just at config time?
An operational boundary is the exact set of operations, resources, systems, and audiences an agent is allowed to touch, the concrete form of least privilege for that agent. You enforce it at runtime by routing every agent request through Oathera Agent Enforcement, which verifies the agent's identity and checks the operation before it runs, and by confining filesystem and network access with the sandboxing tools you already run. A config-file declaration only describes intended behavior; runtime enforcement actually prevents out of scope actions.
Why isn't a config file enough to enforce a boundary?
Because nothing checks it after startup. A config read once describes what the agent should do but cannot stop it from doing more if its credentials allow. Runtime enforcement re-checks every request at a point outside the agent.
Where should the boundary be enforced?
Outside the agent: at Oathera Agent Enforcement in front of your resource servers for API calls, and at the runtime sandbox layer for filesystem and network access. Enforcing inside the agent you are trying to constrain is not a real control.