Skip to content

feat(decisiongrounding): cut AsDecided arms over to Rust Core - #25

Merged
tcballard merged 1 commit into
mainfrom
codex/cut-over-decisiongrounding-to-core
Jul 27, 2026
Merged

feat(decisiongrounding): cut AsDecided arms over to Rust Core#25
tcballard merged 1 commit into
mainfrom
codex/cut-over-decisiongrounding-to-core

Conversation

@tcballard

Copy link
Copy Markdown
Contributor

What changed

  • cut the stable rac and rac_snippets experiment arms from the retired rac executable to AsDecided Core's Rust decided CLI
  • replace RAC_BIN with DECIDED_BIN and remove Python-engine installation guidance
  • record the installed system under test as AsDecided Core in reports and crossover datasets
  • update live scenario tests, preflight checks, documentation, dashboards, and AutonomousQA wiring
  • add CI smoke runs for both AsDecided grounding arms against the pinned released Rust binary

Why

The benchmark already treated the product under test as an external process, but its adapter and provenance still targeted the retired Python-era executable. This completes the Core cutover without changing the historical arm IDs, preserving longitudinal result comparability while making the product and binary identity explicit.

The structured system_under_test envelope also leaves a cleaner provenance boundary for the benchmark's planned evolution into SWE-DecisionGrounding.

Validation

  • 479 passed, 4 skipped across the full DecisionGrounding suite
  • focused cutover contracts: 39 passed
  • live rac arm against decided 0.23.1: 17/17 cells completed
  • live rac_snippets arm against decided 0.23.1: 17/17 cells completed
  • git diff --check

@tcballard
tcballard marked this pull request as ready for review July 27, 2026 07:17
@tcballard
tcballard merged commit 3660023 into main Jul 27, 2026
17 checks passed
@tcballard
tcballard deleted the codex/cut-over-decisiongrounding-to-core branch July 27, 2026 07:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant