[AAASM-4001] 🔧 (ci): Align Codecov measured scope to SonarCloud#202
Conversation
protoc-generated agent_assembly/proto/** is dropped from SonarCloud analysis via sonar.exclusions, but Codecov still counted its near-zero generated-code coverage. Add it to codecov.yml ignore (alongside the existing types.py) so Codecov measures the same file set as sonar.sources, converging the headline coverage % without touching the shared coverage_*.xml or Sonar's Quality Gate. Closes AAASM-4002 Refs AAASM-4001, AAASM-3999
|
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
|
Claude Code — PR review CI state: GREEN. Scope verdict: MATCHES. The diff is exactly what AAASM-4001 specifies — add Side-effect assessment: SAFE. Single-file, config-only diff ( Ready to merge? YES — CI fully green, diff matches the ticket precisely, no side effects on the shared coverage artifact or other consumers. Only the standard ≥1 Pioneer approval remains (I do not approve or merge). |



Description
Adds
agent_assembly/proto/**tocodecov.ymlignoreso Codecov measures thesame file set as SonarCloud's
sonar.sources, closing the residual ~2pt headlinecoverage gap under Epic AAASM-3999.
SonarCloud drops the protoc-generated proto tree from analysis via
sonar.exclusions=agent_assembly/proto/**, but Codecov was still counting itsnear-zero generated-code coverage (
*_pb2.py/*_pb2_grpc.py, 0–58%), draggingthe Codecov number ~1.5pt below Sonar. The existing
types.pyignore (AAASM-3962,mirroring
sonar.coverage.exclusions) is preserved.Measured locally (717 passed,
--cov=agent_assembly):This is a Codecov-only view change:
.coveragercand the sharedcoverage_*.xmlartifact (also read by SonarCloud/Coveralls) are untouched, so no real coverage of
hand-written code is altered and SonarCloud's Quality Gate is unaffected.
Type of Change
Breaking Changes
Related Issues
Testing
pytest test/ --cov=agent_assemblyreproduces the pre/post denominators above;codecov.ymlvalidated as parseable YAML.Checklist
Expected post-fix
Codecov moves 92.97% → ~94.56%, converging with SonarCloud 95.0% within a small
documented tolerance (~0.4pt residual is Sonar's own line/branch coverage engine,
not a scope difference).
Closes AAASM-4001
🤖 Generated with Claude Code
https://claude.ai/code/session_01UTHd8q87SpWGe5he9xFZWN