KCC 10.2

Workslop

Content that passes automated quality gates but provides no real value — generated in seconds, costing colleagues hours to evaluate.

Special ConcernsWorkslopInspector PipelineReview qualityCalibration
Created 2026-06-08 · v0.4.0

Description

Workslop is content that passes automated quality gates but provides no real value — typically generated by an agent and accepted by a human without substantive review. The phenomenon was named by Jamil Zaki (HBR, 2026): outputs created in seconds that cost colleagues hours to evaluate.

FailureEffect
Wastes colleagues' timeThey must evaluate hollow outputs
Degrades trust in agent outputsReal value gets dismissed because of the noise
Makes productivity metrics misleadingOutput volume is up, real value is flat

How KCC Detects It

The Inspector Pipeline includes explicit workslop detection. Signatures include: review velocity anomaly (outputs accepted in less time than reading would require — a 500-line spec approved in 30 seconds); approval rate anomaly (reviewers with rubber-stamp rates over 95%); downstream rework signal (outputs flagged downstream above baseline); confidence calibration drift (agents claiming 0.9 confidence but rejected 40% of the time); and engagement signal (outputs passed forward with no human modification).

Detection Flow

Detected signatures are surfaced to the Inspector's review stage. Maintainers determine the cause:

Likely CauseResolution
The agentCapability-level fix needed
The reviewerCognitive load problem (see Bypass and Human Attention)
The policyGating too lenient
The reviewer's incentivesOrganizational signal needs adjustment

Why Naming It Matters

Many organizations experience workslop without a name for it — engineers feel something is wrong but lack the vocabulary to escalate. Naming and detecting it structurally gives a precise term, permits measurement, enables targeted intervention, and distinguishes workslop from honest failures (a different problem).

Workslop is not bad agents. It is good-looking agents whose output is hollow. The structural fix is to detect the hollowness, not to demand the output look better.