JaCoCo coverage policy

Numbers (this toolkit's defaults)

Metric	Hard floor	Target	New-code rule
Line coverage	90%	95–100%	95% (new + changed lines only)
Branch coverage	90%	95–100%	95%

"Hard floor" = build fails. "Target" = aspirational; not enforced but tracked. "New code" = lines added or modified vs origin/main (PR base).

Maven config

<plugin>
    <groupId>org.jacoco</groupId>
    <artifactId>jacoco-maven-plugin</artifactId>
    <version>0.8.12</version>
    <executions>
        <execution>
            <id>prepare</id>
            <goals><goal>prepare-agent</goal></goals>
        </execution>
        <execution>
            <id>report</id>
            <phase>verify</phase>
            <goals><goal>report</goal></goals>
        </execution>
        <execution>
            <id>check</id>
            <phase>verify</phase>
            <goals><goal>check</goal></goals>
            <configuration>
                <rules>
                    <rule>
                        <element>BUNDLE</element>
                        <limits>
                            <limit><counter>LINE</counter><value>COVEREDRATIO</value><minimum>0.90</minimum></limit>
                            <limit><counter>BRANCH</counter><value>COVEREDRATIO</value><minimum>0.90</minimum></limit>
                        </limits>
                    </rule>
                    <rule>
                        <element>PACKAGE</element>
                        <limits>
                            <limit><counter>LINE</counter><value>COVEREDRATIO</value><minimum>0.85</minimum></limit>
                        </limits>
                    </rule>
                </rules>
            </configuration>
        </execution>
    </executions>
</plugin>

Excludes (the only allowed)

• **/*Application.class (the main class)
• Generated code: **/generated/**, **/openapi/**
• DTOs that are pure records with no logic (when measuring branch coverage; line coverage stays)
• **/config/** is not excluded. Configuration is real code.

Brownfield onboarding

If current coverage is below 90%:

1. Measure current values per package.

2. Write them into .specs/_baseline.json:

   {
     "jacoco": {
       "overall_line": 0.74,
       "overall_branch": 0.62,
       "per_package": { "com.example.shop.checkout": { "line": 0.81, "branch": 0.70 } }
     }
   }
   

3. Set Maven thresholds to current values minus 1% (ratchet).

4. Each feature must improve the metric (or hold) for touched packages.

5. New code is held to 95% regardless of baseline.

"New code at 95%"

The harness computes new-code coverage by intersecting the lines added or changed since git merge-base origin/main HEAD (including uncommitted and untracked files) with JaCoCo's <line nr="N" mi="0" ci="3" mb="0" cb="2"/> entries. A diff line with ci > 0 is covered; one with ci == 0 is uncovered; a diff line with no <line> entry is not executable and is ignored. Threshold: 95% (NEW_CODE_THRESHOLD).

This is implemented in .github/scripts/check-new-code-coverage.sh.

Anti-patterns

• Excluding a class because it's "hard to test".
• <excludes>**/*Service*</excludes> — services are exactly what you must test.
• Treating coverage as the goal. It's a floor, not a target. Mutation testing is the real signal.