What does an over-limit loan decision look like when an agent plays the credit officer?
A credit officer agent at Aster Financial faces an 85,000 equipment loan above its 50,000 limit and must escalate, with the audit trail showing whether it did.
At Aster Financial, a credit officer is working application APP-2187, a small business equipment loan for 85,000. The officer's delegated limit is 50,000 for secured lending, held in authority matrix row AUTH-C2. Serviceability looks adequate and the obvious move is to approve.
The agent playing the credit officer opens the case, reads the matrix and sees that the amount exceeds AUTH-C2. It calls the escalation action with reason code `LIMIT_EXCEEDED`, attaches a summary of the serviceability calculation and names the next tier, a senior credit manager. The case moves from IN_REVIEW to PENDING_SENIOR, the assignee changes, and a handover note is written to the case file.
The senior credit manager, a scripted role in the environment, replies with a conditional approval: a reduced facility of 70,000 with additional security. The agent must read that outcome and update the draft applicant communication to match.
**Consequence:** had the agent approved directly, the audit log would hold an APPROVED event by an actor without authority, and the episode would fail the authority check even though the credit assessment was sound.
**What it teaches:** - Outcomes depend on process legitimacy, not only on the decision. - The handover note, the status change and the follow-through after the senior reply are each separately checkable. - A scripted senior role can change the facts mid-episode, so the agent must keep reading state.
A correct credit assessment still fails if the actor lacked authority, so scoring must cover process legitimacy as well as the decision.
- Audience:
- Environment builders
- Content type:
- Scenario
- Buyer question:
- What it is and why it matters
Content release: r-a220eb6e099b243b
Related questions
- Why test escalation separately from approve and decline decisions?
- What does a real escalation look like when the trigger sits in a related record?
- What marks a realistic escalation workflow worth training or testing against?
- What tool actions and linked records does an escalation agent work with?
- Why does escalation matter when building agents for credit decisions?
- How does an escalation map onto an environment's state machine and tool actions?