Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.
TypeScript ai-sdkclaude-codeclicode-searchcodexcoding-agentscontext-retrievaldeveloper-toolsjevsemantic-searchtypescriptvercel-ai-gateway
docs: pin Django failure to same-type collation case
David Zhangc054f7fba4b7
diff --git a/specs/jevgrep/assets/freshness-confirmation.md b/specs/jevgrep/assets/freshness-confirmation.mdindex ec9c4fd..669dd8c 100644--- a/specs/jevgrep/assets/freshness-confirmation.md+++ b/specs/jevgrep/assets/freshness-confirmation.md@@ -32,3 +32,19 @@ Root execution log: `/tmp/jg-freshness-v1-run.log`. Per-task grade and accountinreceipts, complete event streams, patches and request/response evidence remainwith the attempts. Acceptance stays unproven until all ten outcomes are retainedand the full aggregate is evaluated.++Django's packet contained 51 files in 8,071 bytes without excerpts. SQLite was+listed and its critical type-only rebuild guard was read, untruncated, before the+first patch. The final patch left that guard unchanged. Official evaluation ran+123 tests: 121 passed, one failed and one skipped. The failure is a normal foreign+key retaining no collation after its referenced primary key changes collation+without changing type; later many-to-many and reversal assertions were not reached.++The agent's own regression changed AutoField to CharField, missing that same-type+condition. Its broad suite ran 312 tests: 283 passed and 29 skipped. A separate+hidden-relation fixture failed locally and was simplified without a production+fix, but that is not the official failing case. Paired evidence and the exact+pre-patch SQLite read are retained in+`/tmp/jg-freshness-v1-django-comparison/`. The needed code was delivered and read;+this does not prove model variance alone caused the outcome, nor that adding+initial excerpts would prevent it.