{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"objective_weighting_governance__computer_science","search_lanes":{"direct_problem_and_intervention":{"queries":["continuous integration test scheduling weighted multi-objective runtime failure history change relevance","CI test priorititization weighted score explain ranking security tests","governance sensitivity analysis weights test scheduling"],"source_ids":["SRC1","SRC3"],"no_result_note":null},"synonyms_and_historical_terms":{"queries":["regression test selection resource constrained environments historical test performance","system test case prioritization value driven PORT factor weights","priority merger nightly regression suite omitted tests"],"source_ids":["SRC1","SRC2"],"no_result_note":null},"products_practices_and_standards":{"queries":["TestRail AI test prioritization security label booster explanation ranking","GitHub protected branches required status checks merge documentation","mandatory CI tests branch protection optional test prioritization"],"source_ids":["SRC3","SRC4"],"no_result_note":null},"component_combination":{"queries":["test prioritization stakeholder assigned weights sensitivity analysis","weighted regression test prioritization complaints weight changes consensus","CI scheduler protected mandatory checks weight sensitivity audit trail authorization"],"source_ids":["SRC1","SRC2","SRC3","SRC4"],"no_result_note":"No retained source specifies the complete combination of formally authorized weights, protected non-compensable suites, per-queue contribution traces, predeclared sensitivity alternatives, stakeholder sign-offs, and revision triggers."}},"sources":[{"source_id":"SRC1","title":"Experience Report: Automated System Level Regression Test Prioritization Using Multiple Factors","publisher":"IEEE / Mälardalen University","url":"https://www.es.mdu.se/pdf_publications/4511.pdf","source_type":"PRIMARY_RESEARCH","claims_supported":["An industrial CI-related system used a weighted merger of multiple test priorities, including duration, recent failures, time since execution, and source-code changes.","Limited test time, late fault detection, and repeatedly omitted tests were observed operational problems.","Priority-merger weights were changed after users questioned omissions, and the published report exposes factor contributions and numerical weights.","The scheduler selected tests greedily until a predetermined duration was exhausted, demonstrating consequential weighted omissions under constrained capacity."]},{"source_id":"SRC2","title":"Towards the Prioritization of System Test Cases","publisher":"North Carolina State University Repository","url":"https://repository.lib.ncsu.edu/bitstreams/255fd21a-2b41-4521-91a7-0559fda632ea/download","source_type":"PRIMARY_RESEARCH","claims_supported":["The PORT method assigns values and weights to multiple factors—including customer priority, volatility, complexity, and fault proneness—to prioritize system tests.","The research explicitly examines whether alternative factor-weight allocations can sway prioritization and misguide engineers.","Its sensitivity analysis compares team-assigned, equal, mean-based, variance-based, and contradictory weighting strategies, establishing weight robustness as prior diagnostic practice."]},{"source_id":"SRC3","title":"Introducing TestRail 10.5: AI Test Prioritization","publisher":"TestRail","url":"https://www.testrail.com/blog/introducing-testrail-10-5-ai-test-prioritization/","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["A commercial product prioritizes tests when suites exceed available release time using failures, defects, flakiness, semantic context, and impacted components.","Users can boost security-labeled or other context-specific tests and integrate the resulting rankings into CI/CD through an API.","The product supplies an explanation for every ranking and permits manual overrides, making explanatory traces and contestability established product features."]},{"source_id":"SRC4","title":"About protected branches","publisher":"GitHub Docs","url":"https://docs.github.com/en/repositories/configuring-branches-and-merges/managing-protected-branches/about-protected-branches","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["Repository administrators or authorized custom roles can configure branch-protection requirements.","Required status checks must reach an accepted status before a protected branch can be merged.","Rules can restrict bypasses and require checks from an intended application, providing established mechanisms for non-compensable merge safeguards and defined authority."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The core operational problem is visible: industrial research reports insufficient time, weighted multi-factor prioritization, deadline-based suite truncation, repeated omissions, user questions about omitted tests, and subsequent weight rebalancing. A current product likewise combines historical, flakiness, contextual, and user-supplied priority signals and explains rankings. GitHub demonstrates that some checks can instead be mandatory. The retained evidence does not establish the prevalence of opaque coefficients, organization-wide authority disputes, security suites losing priority in practice, or schedule reversals under modest defensible weight changes.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"closest_prior_art":[{"name":"SuiteBuilder weighted priority merger","source_ids":["SRC1"],"overlap":"A production industrial test scheduler combines several independently generated priorities with explicit numerical weights, selects tests within a time limit, encounters contested omissions, and revises the weight balance in response.","remaining_difference":"It does not describe formal policy authority, security and reliability sign-off, protected non-compensable suites, predeclared sensitivity sweeps, per-queue impact reports, expiring approvals, or revision triggers."},{"name":"PORT value-driven prioritization and weight-sensitivity analysis","source_ids":["SRC2"],"overlap":"Treats test-priority factors as weighted value inputs and evaluates how alternative allocations affect prioritization, including concern that weights may misguide engineers.","remaining_difference":"It is a research method for system-test prioritization rather than governance of live pre-merge CI compute; it lacks protected checks, approval authority, decision-impact traces, stakeholder objections, and a versioned policy lifecycle."},{"name":"TestRail AI Test Prioritization","source_ids":["SRC3"],"overlap":"Combines several test-value signals, supports contextual security boosting, explains individual rankings, permits overrides, and exposes rankings for CI/CD integration under limited testing time.","remaining_difference":"It does not publish stable normalized coefficients or require their organizational authorization, sensitivity replay, assurance sign-off, non-compensability, or revision governance."},{"name":"GitHub protected branches and required status checks","source_ids":["SRC4"],"overlap":"Separates required merge checks from preferences, constrains bypass authority, and can require an intended check provider.","remaining_difference":"It does not govern the composite scoring of optional tests or provide weight disclosure, ranking sensitivity, impact tracing, or approval and revision records."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"For constrained pre-merge CI, coefficients for optional-test scheduling should be governed as a versioned, authorized policy: mandatory assurance checks remain outside compensatory scoring, and disclosed score contributions plus predeclared plausible-weight replays must inform stakeholder approval and later revision. The retained prior art establishes weighted scheduling, contested reweighting, sensitivity analysis, ranking explanations, overrides, security boosting, and mandatory checks separately, but not this combined governance rule.","contrastive_claim_falsifier":"The contrastive claim would be falsified by an earlier operative product, practice, standard, or study that jointly requires authorized and versioned objective weights, non-compensable required suites, per-queue score-impact traces, predeclared weight sensitivity, stakeholder sign-off, and fragility-based approval or revision; it would also fail empirically if bounded replay finds no consequential multi-objective score, deadline omissions, plausible-weight reversals, or contested authority or compensability issue.","gates":{"adequate_source_search":{"status":"PASS","rationale":"The search covered direct proposal language, older regression-test and value-driven terminology, industrial and commercial implementations, required-check practices, weight sensitivity, explanations, and component combinations. Exactly four opened direct sources from four publisher organizations were retained, including two primary studies and two first-party sources.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"Industrial and product evidence directly supports constrained test execution, composite prioritization, consequential omissions, explanations, and weight-related contestation. Evidence for the proposal's broader opacity and authority assertions is incomplete, so the problem is only partly supported.","source_ids":["SRC1","SRC3","SRC4"]},"distinct_testable_claim":{"status":"PASS","rationale":"The remaining claim is distinguishable from a new prioritization algorithm: it requires a specific governance bundle linking formal authority, non-compensability, disclosed impacts, sensitivity fragility, approval, and revision. A closer operative bundle or a replay showing no consequential fragility would falsify it.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"bounded_next_test":{"status":"PASS","rationale":"A read-only replay of one completed seven-day queue for one repository group is bounded and technically grounded in prior uses of historical test data and weight comparisons. It can compare three independently supplied, predeclared weight sets and measure ranking reversals, deadline omissions, contribution changes, data gaps, and protected-check violations without altering production.","source_ids":["SRC1","SRC2","SRC3"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first test is offline, uses existing authorized logs, and changes neither live scheduling nor merge eligibility. Required-check configuration and authority can remain untouched. Explicit halts for unauthorized data, secrets, personnel information, unreconstructable proxies, or unidentified protected suites address the obvious risks; deployment would require separate approval.","source_ids":["SRC3","SRC4"]}},"screen_survival":true,"world_novelty_boundary":"This bounded four-source screen establishes only coarse researchability and adjacent prior art. It cannot establish world novelty, patentability, freedom to operate, market size, expert acceptance, prevalence, or realized value. Unsearched patents, proprietary CI schedulers, internal engineering policies, standards, and differently termed literature may contain closer matches."}