chore: CLAUDE.md proposals for 5 repos with recent merged PRs (May–Jun 2026)

[shrivastavakapil2000]

Summary

Reviews merged PRs from May–June 2026 by team members SayaliPat, shrivastavakapil2000, JoeVsVolcano, mike-brant, and nathan-resonate across the organization, and produces CLAUDE.md files capturing current architecture, key concepts, and recent changes.

Since this session only has write access to resonate/.github, the files are staged here under claude-md-updates/. Each needs to be copied to the root of its target repo and a separate PR opened there.

Files to Apply

File in this PR	Target Repo	Action
claude-md-updates/step-function-workflow-orchestrator/CLAUDE.md	resonate/step-function-workflow-orchestrator	Create
claude-md-updates/batch-expression-modeling/CLAUDE.md	resonate/batch-expression-modeling	Replace existing
claude-md-updates/identity-graph/CLAUDE.md	resonate/identity-graph	Create
claude-md-updates/batch-audience-delivery-syndication/CLAUDE.md	resonate/batch-audience-delivery-syndication	Create
claude-md-updates/dos-data-pipeline/CLAUDE.md	resonate/dos-data-pipeline	Create

What's Documented in Each File

step-function-workflow-orchestrator (NEW)

• 12 active pipeline inventory with purposes
• EMR 5→7 migration notes (Spark 2→3, YARN vcore fix, memory tuning)
• Integration test patterns (pytest, synthetic golden data, 90-min timeout)
• Dynamic dates lambda: directory mode vs flat-file sentinel mode
• Environment table (dev/qa/integration/nonprod/prod)
• Recent: fusion-behavior-preprocess removed, geo district namespace fixes, L2 TapAd stitch wiring

batch-expression-modeling (UPDATE)

• All existing content preserved
• Added: formatter metrics lambda (batch-delivery-formatter-publish-metrics)
• Added: stitch throttle fix — MaxConcurrency=2 + index-staggered Wait (CDP-118972)
• Added: delta_with_full_fallback refresh type requirement in _annotate_previous_date_partitions
• Added: Formatter output path partition order (vendor=*/method=*/)

identity-graph (NEW)

• 11 Spark pipeline jobs with purposes and bug-fix history
• Shared utilities (HashUtils, StagingWriter, AddressNormalizer, IpFilter, ScoringConfig)
• PRISM design overview (6 tracks, 20 Jira tickets)
• All jobs use scopt CLI args (no application.conf)
• Scala 2.12.17 / Spark 3.5.0, JDK 17 --add-opens flags

batch-audience-delivery-syndication (NEW)

• Vendor table (OpenX, Experian, Viant, BlockGraph)
• openx-publish-data-files: hardcoded .csv.gz extension — do not revert (CDP-118955 fix)
• blockgraph-create-taxonomy-file: taxonomy generation, SPI=N constant, initial vs. refresh routing
• Source path partition order: vendor=*/method=av/ (not legacy method=av/vendor=*/)

dos-data-pipeline (NEW)

• district_source provenance enum (L2_CONFIRMED, L2_UNCONFIRMED, IP_INFERRED)
• IP-inferred district fallback via 4 ZIP→district CSVs
• ToBitmap gating on L2_CONFIRMED
• ZIP→district namespace requirements (L2 canonical vs floterial — critical for aggregation joins)
• GeoLocationFullBackfill always re-derives all 4 districts

How to Apply to Target Repos

# For each target repo:
git checkout -b chore/add-claude-md
cp claude-md-updates/<repo-name>/CLAUDE.md ./CLAUDE.md
git add CLAUDE.md && git commit -m "chore: add CLAUDE.md with project guidance for Claude Code"
git push -u origin chore/add-claude-md
gh pr create --title "chore: add CLAUDE.md" --body "Adds Claude Code project guidance based on recent merged PRs."

---

Generated by Claude Code

[Copilot]

Pull request overview

This PR stages CLAUDE.md documentation updates (generated from recent May–June 2026 merged PR activity) for five other repositories under claude-md-updates/, along with a local README describing how to copy/apply them in their target repos.

Changes:

• Added new CLAUDE.md drafts for: step-function-workflow-orchestrator, identity-graph, batch-audience-delivery-syndication, and dos-data-pipeline.
• Replaced/updated the CLAUDE.md draft for batch-expression-modeling.
• Added claude-md-updates/README.md with copy/apply instructions.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
claude-md-updates/README.md	Documents which staged CLAUDE.md maps to which target repo and how to copy/apply them.
claude-md-updates/step-function-workflow-orchestrator/CLAUDE.md	Draft guidance for Step Functions + EMR pipeline orchestrator repo (pipelines, EMR7 migration notes, testing patterns).
claude-md-updates/batch-expression-modeling/CLAUDE.md	Draft guidance for BEM system (architecture, workflows, JSONata notes, recent operational fixes).
claude-md-updates/identity-graph/CLAUDE.md	Draft guidance for Identity Graph / PRISM Spark jobs and build/deploy/test workflows.
claude-md-updates/batch-audience-delivery-syndication/CLAUDE.md	Draft guidance for syndication workflows/lambdas and vendor-specific gotchas.
claude-md-updates/dos-data-pipeline/CLAUDE.md	Draft guidance for DOS geo/district pipeline concepts and recent changes.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.