Escalation is a successful outcome
An agent that recognizes it has exceeded its contract is safer than one that guesses. The goal is a visible decision, not uninterrupted automation.
Answer page
When should a human stop an agent and escalate the work?
Escalate when scope, product intent, risk, or evidence can no longer be resolved within the task's declared boundary.
Direct answer
Stop and escalate an agent when it encounters ambiguous requirements, needs out-of-scope changes, cannot run the required validation, finds conflicting evidence, or reaches a security, architectural, or customer-impacting decision that needs accountable human judgment.
Practical guidance
Make the next review decision easier.
Keep a narrow trigger
Define escalation triggers in the task: new target paths, missing validation, unclear acceptance, secrets or permissions, and conflicting architectural constraints.
Turn the answer into a new boundary
After a human resolves the issue, update the task or create a follow-up task so the next agent does not rediscover the same ambiguity.
Verified demo evidence
A public CLI result, not a completion claim.
This command and result are from the website’s checked-in synthetic repository demo. Substitute your own repository paths and validation command when you apply the workflow.
$ day-shift init
Created .day-shift workspace
Next: run day-shift doctorAuthorship and sources
Trace this guidance to maintained product evidence.
- Maintainer
- Tianna McCoy ↗Day Shift maintainer; responsible for the repository-native workflow and release evidence referenced here.
- Last updated
- Tested Day Shift
- v0.1.24
Keep exploring
Follow the next question, not a generic funnel.
What Should a Coding-Agent Task Definition Include?
The objective, scope, acceptance criteria, and validation a coding agent needs before changing code.
Read nextCoding Agents on Regulated or Security-Sensitive Codebases
Use coding agents within least-privilege, review, validation, and approval controls.
Read nextHow to Recover When a Coding Agent Makes the Wrong Change
Contain, review, correct, and improve the task boundary after an agent mistake.
Read nextFind your answer
Have we answered your question?
Search the answer pages, concept, comparisons, demo, and FAQ without leaving the site.
Search by a problem, workflow step, or tool boundary.