Rendered from docs/process/obligations/0239-three-run-tooling-gaps-from-run-20261003-1026-filed-together.md in the Headwater corpus. Every document on this half of the site is typed by the taxonomy the descriptor names: corpus.json.

Project process. These notes record how Headwater itself is built and run. You do not need them to use Headwater. Back to the specification

Three run-tooling gaps from run 20261003-1026, filed together

Context

hw-run-policy sends a finding about the build order, its tooling or the CI of this repository to this shelf. Run 20261003-1026 wrote these as intake lines 3, 4 and 27, from the work on #1670. The product owner ruled each one RECORD at the end-of-run pass on 2026-10-09. Each item was checked on 66dec0f9. The product owner read the steps of the headwater job in ci.yml, the traces_to list of the blind-grade evaluation, and cmd_grade in tools/run/relevance-grade.sh.

Obligation

  • HW-OBL-0145 says "The job carries 55 steps". The headwater job now carries 89 entries in its steps list, so the count and the split that follows it are stale.
  • The blind-grade evaluation, docs/evaluations/what-a-blind-grade-of-the-two-routing-paths-offers-showed.md, now lists tools/run/relevance-grade.sh under traces_to. No document names tools/run/relevance-grade-fixtures.sh or the record under tools/run/relevance-grade/, so both stay in the gap that HW-OBL-0233 names.
  • Mutant N5 of cmd_grade survives: removing the test that refuses a sample row graded twice leaves the suite green. A case that appends one sample row twice and expects exit 4 with no grade.txt would hold it. #1689 changes the same suite, so it is the natural change to carry this case.

Discharge

Each item discharges alone. The first does when HW-OBL-0145 states the count the job holds or drops it. The second does when a document names the fixtures and the record, and the third when a case kills N5. The record discharges when every item has.