feat(data-quality): add W65 durable lifecycle store

[ammarheidari]

Authority

W65 / #214 Slice 4A is protected-main admitted via PR #252 at baseline 9572726f9880e82aaf1b836b0a8c1e330c5ca568.

Slice 4B scope

• provider-neutral ADO lifecycle/evidence store;
• SQLite standalone connection factory with WAL/foreign-key/busy-timeout settings;
• PostgreSQL connection factory with HA-capable migration locking;
• schema versioning under kafdeck_schema_info;
• create vs revision-guarded replace/state update semantics;
• stable bounded policy pagination;
• idempotent aggregate/progress evidence append;
• bounded evidence time-range queries;
• SQLite round-trip/idempotency/pagination tests.

Safety

• no Kafka reads or consumer-group membership;
• no offset commits;
• no raw key/value/header/payload persistence;
• durable evidence remains aggregate/progress only;
• no unbounded scans: policy and evidence queries retain contract ceilings;
• no worker/API/UI activation in this slice.

Final corrective exact head: db3feabe59f8b4139b5dcf4900afe9094ceae952.

Refs #214 #209.

Corrective closure

• policy scope is part of durable evidence identity;
• state transitions use atomic CAS with UPDATE ... RETURNING;
• evidence queries use a complete deterministic identity ordering;
• shared kafdeck_schema_info creation is serialized with the same advisory lock as historical metrics;
• policy persistence now uses an explicit persistence DTO rather than deserializing domain constructors directly.
• historical metric maintenance now uses the same shared schema advisory lock, closing the remaining startup DDL race.
• mutation operation, fleet mutation state, and connect auto-restart initializers now also join the shared kafdeck_schema_info advisory-lock protocol.
• corrected the fleet initializer so shared schema transaction scope is confined to InitializeAsync; no schema transaction leaks into fleet conflict backfill.

[ammarheidari]

@codex review

Please review exact head ccbcc033559df9a77bf324ae357dfffe2dbc19ad only. Focus on SQLite/PostgreSQL provider-neutral SQL, optimistic revision correctness, stable pagination, idempotent evidence identity, schema initialization concurrency, and proving no raw Kafka material can become durable.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-04T19:43:19.037749Z	db3feab	Manual request

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ccbcc03355

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[ammarheidari]

@codex review

Please review corrective exact head 24d8958783ea8e32dfceba2e410d25c9d65cf9bf only. Verify the prior four findings and CI serialization failure are closed: scope-qualified evidence identity, atomic state CAS returning the winning definition, deterministic evidence ordering, shared schema migration lock, and explicit persistence DTO round-trip.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 24d8958783

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[ammarheidari]

@codex review

Please review final corrective exact head 81407f9c9771bbfae8cd70909dcca8379107d584 only. Verify all prior findings are closed, including the remaining shared-schema startup race: data-quality, historical-metric, and historical-metric-maintenance initializers now serialize kafdeck_schema_info creation with the same advisory lock.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 81407f9c97

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[ammarheidari]

@codex review

Please review final exact head 5ae2c218963fea7c3faf6673fe73fc1e1d2e06a7 only. Verify all prior findings are closed and every PostgreSQL initializer that creates kafdeck_schema_info participates in the same shared advisory-lock protocol, including mutation operations, fleet mutation state, and connect auto-restart.

[ammarheidari]

@codex review

Please review final exact head db3feabe59f8b4139b5dcf4900afe9094ceae952 only. Verify all previous persistence/concurrency findings remain closed, all shared-schema creators join the common advisory-lock protocol, and the fleet initializer compile regression from the prior head is fully corrected without leaking schema transaction scope into conflict backfill.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Nice work!

Reviewed commit: db3feabe59

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".