KB-A6C1
Main FF After DDL Ratify · 02 Test + Static Result
4 min read Revision 1
dot-iu-cutterv0.5main-fast-forward-after-lifecycle-ddl-ratifytests-staticg2-passdieu442026-05-20
dot-iu-cutter v0.5 — Main FF After DDL Ratify · Test + Static Result
doc 2 of 5 · 2026-05-20
phase : G2 — tests + static checks before FF outcome : PASS (264/265 regression, 1 known baseline; 4/4 fingerprint byte-equality) production_mutation : NONE (no DB connection from tests)
1. Regression suite
command : python3 -m pytest tests/ -q
result : 264 passed, 1 failed in 0.26s
new_regression_caused_by_this_macro : false (this macro has NOT changed code)
Single failure detail (pre-existing baseline R-8)
test : tests/test_security_boundaries.py::TestNoSecretPrinted
::test_source_has_no_hardcoded_dsn_or_secret
target_file : cutter_agent/cutwrite.py
flagged : 'PGPASSWORD'
context : inside DB_ENV_GUARD tuple as an env-var NAME guard
(the writer REFUSES to run if PGPASSWORD is set;
the string is a guard-list entry, NOT a secret)
pre_existing : true
documented_in:
- v0.5-post-enactment-closeout-release-readiness/05-remaining-risk-and-backlog.md (R-8)
- v0.5-lifecycle-ddl-ratification-runbook/04-test-static-check-result-2026-05-20.md
backlog_item : B-DB-ENV-GUARD-BASELINE-CLEAN (LOW; deferred)
caused_by_this_macro : false (no .py file modified)
The same failure was present at every recent HEAD (32cfa93 ← 7133c44 ← 6a56bc3 ← … ← 4367c83 baseline). Fast-forwarding main from 32cfa93 to 1cd286e does NOT introduce it; both points already carry it.
2. sql/lifecycle/ inventory
README.md 5,431
bundle_a_vocab_and_log.sql 4,855
bundle_b_immutability.sql 6,867
bundle_c_fn_iu_enact.sql 11,571
bundle_d_gateway_and_grants.sql 3,535
bundle_e_fn_iu_apply_edit_draft_patch.sql 7,104
fingerprints.yaml 5,856
rollback_runbook.sql 7,642
verify_behavioral_probes.sql 3,900
verify_postapply.sql 7,376
verify_preflight.sql 3,016
------
11 files 67,153 B
All 11 files documented in the prior ratification closeout are present and unmodified since commit 1cd286e.
3. Function body fingerprint check (md5 byte-equality vs live pg_proc.prosrc)
function len md5(extracted) verdict
-----------------------------+------+----------------------------------+--------
fn_iu_enacted_immut 1631 aeb3fa4fdb225f6ba6b7073582caa454 OK
fn_uv_enacted_immut 2415 03f035a23cbc79a9e811a6da6f5266ba OK
fn_iu_enact 8674 6ca9bc39e2d2be93dd8a71739fa80dc4 OK
fn_iu_apply_edit_draft 4826 42e96b6c9e81a2d0a28b30644d178a26 OK
Extraction uses regex AS \$tag\$(.*?)\$tag\$\s*; on the committed bundle SQL files (B/C/E). All four function bodies remain byte-equal to the live PG prosrc.
4. No production touch this gate
PG_writes : NONE
PG_reads : NONE (this macro reads ONLY git + KB + filesystem;
no PG connection issued)
file_writes_outside_repo : NONE
repo_writes : NONE this gate (writes happen in G3)
git_writes : NONE this gate
deploy/restart/push/tag : NONE
5. Gate disposition
G2_regression_no_new_failures : PASS (R-8 baseline, pre-existing)
G2_sql_lifecycle_artifacts : PASS (11/11 files present)
G2_fingerprint_byte_equality : PASS (4/4 functions)
G2_no_PG_or_deploy_touched : CONFIRMED
6. STOP
Tests + static checks pass. Proceed to doc 03 (FF execution log).