Add pipeline prefixes to S3 bucket for test/synthetic eCRs

[nickclyde]

Description

Rather than relying on APHL's ecr-data-repository bucket, we'll add pipeline prefixes to the existing dibbs-text-to-code S3 bucket (defined in terraform/s3.tf) for test/synthetic eCRs and pipeline artifacts.

Prefixes needed (mirroring APHL's structure):

• Input: eICR documents, Schematron error files, TTC submission triggers
• Output: TTC output, TTC metadata, augmented eICRs, augmentation metadata

Acceptance Criteria

• Pipeline prefixes added to the existing dibbs-text-to-code bucket
• All prefixes for both lambdas are defined
• S3 event notifications can be attached for SQS triggers
• Bucket policy scoped appropriately for Lambda access
• Test/synthetic eCR data can be uploaded for E2E testing
• Document how prefix structure maps to APHL's ecr-data-repository layout so the transition to production is clear

Notes

The prefix structure should mirror APHL's conventions (TextToCodeValidateSubmissionV2/, AugmentationEICRV2/, etc.) so that switching to their bucket in production is just a configuration change (env vars / TF variables), not a code change.

Depends On

• Update TTC Lambda input and output buckets #349 (prefix finalization)

[nickclyde]

Core work completed across PRs #419 and #424. PR #419 aligned all S3 prefix defaults with APHL conventions and added source-bucket routing. PR #424 adds EventBridge/SQS event notifications for the augmentation lambda. Remaining items (bucket policy scoping, test data tooling) are general concerns not specific to this issue.