Add Tier 2 differentiators: cost metrics, root cause, replay, mutation, fingerprint#66
Merged
pratyush618 merged 7 commits intomainfrom Apr 7, 2026
Merged
Conversation
CostNormalizedMetric, LatencyNormalizedMetric, CostEfficiencyAnalyzer, ParetoFrontier in agenteval-metrics/cost package, 29 tests.
RootCauseAnalyzer clusters regressed cases by failure pattern, detects output/tool/cost/latency changes, ranks by impact, 11 tests.
RecordingJudgeModel/AgentWrapper decorators, ReplayJudgeModel/AgentWrapper for $0 regression tests, RecordingStore persistence, ReplaySuite orchestrator, 32 tests.
Sealed Mutator interface with 5 built-in mutators, PluggableMutator, MutationSuite orchestrator, AgentFactory, 22 tests.
CapabilityDimension enum (8 dimensions), CapabilityProfiler orchestrator, CapabilityComparison, CapabilityReporter, 17 tests.
Update README module structure, add 6 doc pages under docs/advanced for contract testing, chaos engineering, statistical analysis, deterministic replay, mutation testing, and capability fingerprinting.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Five Tier 2 differentiating features — no competing eval tool offers any of these:
agenteval-metrics/cost):CostNormalizedMetric,LatencyNormalizedMetric,CostEfficiencyAnalyzerwith Pareto frontier computation — 29 testsagenteval-reporting/regression/rootcause):RootCauseAnalyzerclusters regressed cases by failure pattern, detects output/tool/cost/latency changes, ranks by impact — 11 testsagenteval-replay): Record agent+judge interactions, replay without API calls ($0 regression tests).RecordingJudgeModel/ReplayJudgeModeldecorators,RecordingStorepersistence,ReplaySuiteorchestrator — 32 testsagenteval-mutation): SealedMutatorinterface with 5 built-in mutators (weaken constraints, remove safety, inject contradiction, etc.),MutationSuiteorchestrator measures eval detection rate — 22 testsagenteval-fingerprint):CapabilityProfilerevaluates agents across 8 dimensions,CapabilityComparisonfor side-by-side profiles,CapabilityReporterfor console output — 17 tests58 new files, ~4,850 lines, 111 new tests — all passing.
Test plan
mvn test -pl agenteval-metrics— cost metrics pass (29 new)mvn test -pl agenteval-reporting— root cause analysis pass (11 new)mvn test -pl agenteval-replay— replay module pass (32 tests)mvn test -pl agenteval-mutation— mutation module pass (22 tests)mvn test -pl agenteval-fingerprint— fingerprint module pass (17 tests)mvn clean install -Denforcer.skip=true