[prathshenoy]

Summary

What:

• Generate request-level issue summaries from the shared issue history.
• Record repeat reports during normal processing and dead-letter recovery.

Why:

• Keep request results consistent with accepted issue updates after failures. Retaining an event and repairing its projections before advancing the cursor lets normal processing and recovery finish interrupted writes without inventing reports.
• Distinguish newly covered requests from historical requests whose issue summaries were never recorded. Recording coverage begins only after all issue-tracking workers use the updated contract.

Test Plan

• Add unit tests.
• Verify replay rebuilds missing summaries against a real database.
• Verify stale workers cannot reopen issues or corrupt request results after recovery and a newer build.

Revert Plan

• Revert dependent request-result reads and this PR. Operational issue state and retained issue history remain available; summary rows can remain unused.

Issue Links

• CODEM-607

Stack

1. feat(stovepipe): Store request issue summaries #1023
2. feat(stovepipe): Retain repeat issue reports #1027
3. feat(stovepipe): Materialize request issue results #1024
4. feat(stovepipe): Add request issue lookups #1025