feat(crud-service): cutover to MAF runtime

[Cataldir]

Problem statement

Migrate the crud-service bounded context from the legacy BaseRetailAgent runtime internals to the MAF-backed runtime introduced in #981. crud-service is a pure microservice with no agent logic, but it binds the new three-tier memory contract (ADR-032, PR #978) and is the first cutover step per the R1 sequencing — establishing the steady-state pattern (eval baselines, AGC canary, MCP-only A2A lint, span contract validation) that every subsequent bounded context follows.

This is a hard sunset: legacy runtime imports are deleted in the same PR. There is no coexistence flag.

Required change

• Replace any direct usage of lib/src/holiday_peak_lib/agents/ internals in apps/crud-service/src with the public surface backed by the MAF runtime from feat(lib): MAF runtime adapter behind the existing BaseRetailAgent surface #981.
• Bind the three-tier memory contract: wire MemoryFacade (Hot=Redis, Warm=Cosmos, Cold=Blob) via env vars REDIS_URL, COSMOS_ACCOUNT_URI, COSMOS_DATABASE, COSMOS_CONTAINER, BLOB_ACCOUNT_URL, BLOB_CONTAINER.
• Confirm REST endpoints (/products, /orders, /cart) remain unchanged in shape; migration is internal.
• Per-service Foundry config (FOUNDRY_AGENT_ID_FAST, MODEL_DEPLOYMENT_NAME_FAST, FOUNDRY_AGENT_ID_RICH, MODEL_DEPLOYMENT_NAME_RICH, FOUNDRY_STREAM, PROJECT_ENDPOINT) wired via Helm values + Key Vault refs (where applicable for downstream agent calls from CRUD).
• Delete any pre-MAF imports left in apps/crud-service/src.

Affected components

• apps/crud-service/src
• apps/crud-service/tests
• apps/crud-service/Dockerfile (if env-var contract changes)
• Helm values for crud-service deployment.

Acceptance criteria (per spec — Acceptance gate per bounded context)

• Contract tests in lib/tests pass.
• All apps/crud-service unit and integration tests in apps/crud-service/tests pass.
• Continuous eval baseline (per ADR-028 final schema, PR Add continuous agent evaluation engine #974) within maxDeltaPercent for any agent calls originating from crud-service (advisory until PR Add continuous agent evaluation engine #974 merges).
• AGC canary 5 → 25 → 50 → 100 succeeds in dev, then in staging (per ADR-029, PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978).
• No A2A HTTP detected by scripts/ci/lint_no_a2a_http.py (per ADR-030, PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978).
• Span attribute contract validated by App Insights workbook query (per ADR-031, PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978).
• No pre-MAF imports remain in apps/crud-service/src.
• Per-service Foundry config wired via Helm values + Key Vault refs.
• 24-hour steady-state hold satisfied before the next context's PR opens.

Risks and dependencies

Risk	Mitigation
Memory contract binding regresses Hot/Warm/Cold tier latencies.	App Insights workbook compares P50/P95/P99 per memory tier across the cutover; tolerance P95 within 10 %. Outside tolerance halts the canary.
Reach-through into BaseRetailAgent internals from crud-service is found late.	Pre-flight grep across apps/crud-service/src for direct use of internal modules; blocks cutover until refactored.
Foundry agent ID provisioning fails for downstream agent calls from CRUD.	Provisioning runs ahead of this PR; failure halts this context.
Hard-sunset rollback restores legacy code from the revert.	No coexistence state allowed (per spec). Either fully on MAF, or fully on legacy.

Blocked by: #981 (lib MAF runtime seam).

Blocks: the next context (truth-*) per the R1 sequencing rule (one bounded context at a time).

Evidence links

• apps/crud-service
• apps/crud-service/src
• apps/crud-service/tests
• apps/crud-service/Dockerfile
• apps/crud-service/README.md
• lib/src/holiday_peak_lib/agents/base_agent.py

ADR impact

• ADR-007 (memory tiers) — refined by ADR-032 (PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978): crud-service adopts the formal Hot/Warm/Cold binding.
• ADR-024 → ADR-030 (PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978): MCP-only A2A enforced.
• ADR-029 (PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978): AGC canary 5→25→50→100 gates the cutover.
• ADR-031 (PR docs(adr): ADR-029..033 - architecture contracts (canary, MCP A2A, OTEL, memory, UI on SWA) #978): span attribute contract validated.

Branch convention

feature/<issue-id>-crud-maf-cutover (per ADR-018).

BPMN process

%%{init: {'theme':'base', 'themeVariables': {
  'primaryColor':'#FFB3BA',
  'primaryTextColor':'#000',
  'primaryBorderColor':'#FF8B94',
  'lineColor':'#BAE1FF',
  'secondaryColor':'#BAE1FF',
  'tertiaryColor':'#FFFFFF'
}}}%%
flowchart LR
  A[Analyze Current Code] --> B[Design Change]
  B --> C[Implement on Issue Branch]
  C --> D[Open PR]
  D --> E[Validation and Fixes]
  E --> F[Merge to Main]
  F --> G[Monitor Workflows]
  G --> H[Close Issue and Cleanup]

Loading