wagoodman/dive · error

evaluation failed

Error message

evaluation failed

What it means

Thrown at the end of a `dive --ci` run when the configured CI rules evaluated against the image analysis did not all pass (eval.Pass == false). It is not an internal failure: the analysis completed, the report was already published via bus.Report, and this error exists purely to make the dive process exit non-zero so CI pipelines fail. The detailed per-rule results (which thresholds were breached) are printed to the report output before the error is returned.

Solutions

1. Read the CI report output above the error to see which rule(s) failed and by how much
2. Fix the Dockerfile: merge RUN steps that add and remove files in the same layer, drop caches (rm -rf /var/lib/apt/lists/*), remove duplicated files across layers
3. If the threshold is intentionally strict for new images, tune the rule values in the CI config file (e.g. .dive-ci) to realistic levels
4. Disable a rule you do not want enforced by setting its config value to the disabled sentinel instead of deleting analysis steps

Example fix

# before
FROM ubuntu
RUN apt-get update && apt-get install -y curl
RUN rm -rf /var/lib/apt/lists/*

# after (add/remove in one layer, no wasted bytes)
FROM ubuntu
RUN apt-get update && apt-get install -y curl && rm -rf /var/lib/apt/lists/*

# pre-check thresholds before enforcing the gate
dive myimage:latest --json /tmp/dive.json
# inspect efficiencyScore / wastedBytes in the JSON, compare to your thresholds, then decide

# in a CI script: treat exit!=0 as gate failure, but read the report first
dive myimage:latest --ci || { echo 'dive CI gate failed — see report above'; exit 1; }

When it happens

Trigger: Running `dive <image> --ci` (or with ci.enabled in config) where at least one of lowestEfficiencyThreshold, highestWastedBytes, or highestUserWastedPercent rules fails: e.g. image efficiency below 0.9, wasted bytes above 1GB, or user-wasted percent above 0.1 with defaults enabled.

Common situations: A Dockerfile regression adds duplicated files across layers (apt caches left behind, COPY of files later deleted), dropping image efficiency below the configured threshold in a CI gate; or a team tightens thresholds in .dive-ci without rebuilding base images that never met them.

Related errors

• failed evaluation
• cannot find docker client executable
• highestUserWastedPercent config value is outside allowed…
• invalid highestUserWastedPercent config value, given
• invalid highestWastedBytes config value, given

AI-assisted analysis of wagoodman/dive@d6c691947f (2026-08-15). Data as JSON: /api/errors/54e68d3b92d792d3. Report an issue: GitHub.

Appendix: source

Thrown at cmd/dive/cli/internal/command/root.go:91

func run(ctx context.Context, opts options.Application, img *image.Image, content image.ContentReader) error {
	analysis, err := adapter.NewAnalyzer().Analyze(ctx, img)
	if err != nil {
		return fmt.Errorf("cannot analyze image: %w", err)
	}

	if opts.Export.JsonPath != "" {
		if err := adapter.NewExporter(afero.NewOsFs()).ExportTo(ctx, analysis, opts.Export.JsonPath); err != nil {
			return fmt.Errorf("cannot export analysis: %w", err)
		}
		return nil
	}

	if opts.CI.Enabled {
		eval := adapter.NewEvaluator(opts.CI.Rules.List).Evaluate(ctx, analysis)

		if !eval.Pass {
			return errors.New("evaluation failed")
		}
		return nil
	}

	bus.ExploreAnalysis(*analysis, content)

	return nil
}

View on GitHub (pinned to d6c691947f)