Skip to content

Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.

docs: retain repeat changes and final release archive identity

David Zhang31746442d4fc
diff --git a/specs/done/jevgrep/assets/release-0.1.0.md b/specs/done/jevgrep/assets/release-0.1.0.md
index 36e5bea..a145ff9 100644
--- a/specs/done/jevgrep/assets/release-0.1.0.md
+++ b/specs/done/jevgrep/assets/release-0.1.0.md
@@ -3,17 +3,23 @@
The user explicitly authorized publishing `0.1.0` after the variance repeat and
closeout finish. Publication is pending; preparing an archive is not publication.
-The local release candidate is
-`/tmp/jg-010-release/dzhng-jevgrep-0.1.0.tgz`, SHA-256
-`a6c9061c1a7102bd192df8f1d0c2c954b470f86db3ed279818bf167660332f00`.
+The final local release candidate is
+`/tmp/jg-010-final-release/dzhng-jevgrep-0.1.0.tgz`, SHA-256
+`447cdeb88695b6851efb7fbe4bfd628a69e1c83b2f1651e7ed3f3453fe747523`.
It passed the release identity/content/license validator. Compared with the
frozen benchmark archive, only package metadata and the bundled executable
changed; the executable difference is exactly `0.0.0` to `0.1.0`. All other
-archive members are byte-identical. Proof: `/tmp/jg-010-version-parity.json`.
+archive members are byte-identical. Proofs: `/tmp/jg-010-version-parity.json`
+and `/tmp/jg-010-final-parity.json`. Final formatting changed dependency-key
+order only; package metadata values and all runtime bytes are unchanged from
+the preceding local `0.1.0` candidate.
-This candidate passed both native macOS arm64 installed journeys on Node 24.14.0
-and all 24 Linux arm64 installed tests. Logs: `/tmp/jg-010-native.log` and
-`/tmp/jg-010-installed.log`. Full deterministic verification of the unchanged
+The preceding local `0.1.0` candidate passed both native macOS arm64 journeys on
+Node 24.14.0 and all 24 Linux arm64 installed tests. The final metadata-formatted
+archive passed native smoke again and three focused Linux installed journeys.
+Logs: `/tmp/jg-010-{native,installed}.log` and
+`/tmp/jg-010-final-{native,installed}.log`. Repository formatting now passes.
+Full deterministic verification of the unchanged
implementation is retained in [freshness verification](source-freshness.md).
The tag workflow still has to run its own full verification, pack and validate
diff --git a/specs/done/jevgrep/assets/variance-repeat.md b/specs/done/jevgrep/assets/variance-repeat.md
index 381388f..f10213f 100644
--- a/specs/done/jevgrep/assets/variance-repeat.md
+++ b/specs/done/jevgrep/assets/variance-repeat.md
@@ -55,3 +55,21 @@ These are observed differences, not proof that ranking caused the miss. Evidence
Astropy solved again. Repeat Sol cost is $0.3507346 versus first-run $0.3914126
and baseline $0.4286310, retaining a successful cost win. Its observed Jev total
is complete at $0.076535382 and remains excluded from scored Sol cost.
+The repeat used ten generations versus fourteen and restored a lazy forward-call
+guard for the no-dropped-axis case. Its authored regression uses celestial and
+spectral coordinates rather than the first run's linear fixture. Final local
+coverage passed 107, skipped seven and retained the known expired-leap-second
+failure. Evidence: `/tmp/jg-repeat-astropy-comparison.md`.
+
+Xarray solved again at $0.5114540 versus first-run $0.3182098 and baseline
+$0.4937066, losing its earlier cost-win classification. The production patch
+matches the first run. Its packet grew from 2,787 bytes without excerpts to
+17,524 bytes with 12 blocks; the agent ran three successful test selections
+instead of one and added default-attribute and groupby coverage. These observed
+changes accompany higher cost without isolating its cause. Jev's known subtotal
+is $0.181858194, full total unknown. Evidence: `/tmp/jg-repeat-xarray-comparison.md`.
+
+Requests solved on the repeat after failing the first corrected run. Its complete
+Sol cost is $0.2721328 versus first-run $0.2388080 and baseline $0.2685004, so the
+recovered solve is not a strict cost win. Jev's complete observed total is
+$0.010286388, excluded from scored cost.