Terms / Execution Quality / Depth Fixation

Depth Fixation

The agent keeps drilling deeper into implementation detail, generating new sub-tasks and refinements rather than recognizing the work is done and returning control to the user. Expands depth rather than breadth.

Depth Fixation is the vertical counterpart to Agentic Scope Creep's horizontal expansion. Instead of adding new features, the agent endlessly refines the current one: optimizing performance that's already acceptable, adding error handling for scenarios that will never occur, refactoring working code to be "cleaner," or generating test cases for edge cases that don't matter. The agent creates a recursive loop of self-generated work, each layer deeper than the last, and it can be surprisingly difficult to get the agent to stop and surface. It's a failure to recognize "done."

Example

A user asks an agent to write a function that validates email addresses. The agent writes the function (correct and working), then decides to add comprehensive RFC 5322 compliance, then adds internationalized email support, then builds a full email validation library with configurable rules, then starts writing documentation for the library, then begins creating a test suite with 47 edge cases including historically valid but currently deprecated formats. The user wanted a 10-line function. The agent is now 30 turns deep into building an email validation framework and shows no signs of stopping.

Depth Fixation

More in Execution Quality

Execution Fidelity

Plan Rot

Plan Drift

Execution Hallucination

Execution Integrity

Plan Resilience

Task Accountability

Agentic Scope Creep