Sub-issue of #299 (artifact-completeness-benchmark, ADR-093).
Corpus: Initiative 1 of decisions/roadmaps/future/artifact-completeness-benchmark.md.
Scope: the make-or-break — assemble targets that have both a recorded AsDecided artifact corpus and executable tests, sequenced by honesty:
- (a)
AsDecided Core dogfoods itself — rebuild a AsDecided Core component from AsDecided Core's own rac/ artifacts, scored by the engine's tests. Available now; brutally honest; N=1.
- (b) OSS retrofit — open-source projects that carry ADRs plus tests.
- (c) synthetic — authored artifact bundles for known components.
Each target ships its artifact bundle, its component boundary, and its executable test set. Built in asdecided/benchmarks (ADR-092), not AsDecided Core.
Tick on #299 when at least the AsDecided Core dogfood target is runnable.
Sub-issue of #299 (
artifact-completeness-benchmark, ADR-093).Corpus: Initiative 1 of
decisions/roadmaps/future/artifact-completeness-benchmark.md.Scope: the make-or-break — assemble targets that have both a recorded AsDecided artifact corpus and executable tests, sequenced by honesty:
AsDecided Coredogfoods itself — rebuild aAsDecided Corecomponent fromAsDecided Core's ownrac/artifacts, scored by the engine's tests. Available now; brutally honest; N=1.Each target ships its artifact bundle, its component boundary, and its executable test set. Built in
asdecided/benchmarks(ADR-092), notAsDecided Core.Tick on #299 when at least the
AsDecided Coredogfood target is runnable.