forked from wangziqi/ruoyi-vue-pro
166 lines
9.0 KiB
Markdown
166 lines
9.0 KiB
Markdown
# Vertical Slice Roadmap
|
|
|
|
> Phase 0 static assessment generated on 2026-07-29. No build, test, application startup, PostgreSQL connection, or Flyway migration was executed during this assessment.
|
|
|
|
Tickets are vertical behaviors, not technical layers. Work blockers first and use a fresh implementation context per ticket.
|
|
|
|
## EDU-P0-S0 — Baseline, completeness, and architecture decision gate
|
|
|
|
- **Outcome:** Verified baseline, capability/API/database/module-reuse maps, commit reviews, decision register, corrected documentation plan, and bounded first-slice specification.
|
|
- **Risk:** High provenance and operational risk; no implementation should begin from an unclassified dirty baseline.
|
|
- **Blockers:**
|
|
- Obtain or explicitly approve source baseline main/033701a because no source feature/education-core-loop ref exists.
|
|
- Record current target status and preserve all 65 modified tracked and 97 untracked entries.
|
|
- Complete the ten GOAL.md Phase 0 artifact dispositions and separate committed, dirty, absent, and runtime-unverified behavior.
|
|
- Decide tenant origin-binding, principal policy, provider authority, option schema, and PUBLIC graph semantics.
|
|
- **Verification:**
|
|
- Read GOAL.md and target/source rules and module documentation.
|
|
- Record both Git statuses, histories, and refs without destructive commands.
|
|
- Review 11e9cc6, 0f846fd, and committed core-loop evidence ce02f8a.
|
|
- Confirm no tests, builds, PostgreSQL connections, Flyway migrations, or runtime flows were executed.
|
|
- Inventory Auth/Profile/extended Learning, granular tenant-admin, platform-admin, worker/scanner, and database object surfaces.
|
|
|
|
## EDU-P0-S1 — Provider-neutral safe question and session restoration
|
|
|
|
- **Outcome:** Both provider modes and restored sessions fail closed for malformed/unavailable published question content while preserving tenant scope and safe projections.
|
|
- **Risk:** Medium implementation risk and high assessment-integrity/security risk if any alternate path remains fail-open.
|
|
- **Blockers:**
|
|
- EDU-P0-S0 provider and option-contract decisions.
|
|
- Existing dirty provider/session files must be separated from unrelated work.
|
|
- Safe response and question-type option semantics must be fixed.
|
|
- **Verification:**
|
|
- Provider contract tests for Scalar and Java.
|
|
- Browsing, collection, practice-create/restore, malformed, unavailable, unpublished, cross-tenant, PUBLIC, and sensitive-field tests.
|
|
- Run focused tests/compile/diff checks only after authorization and report exact results.
|
|
|
|
## EDU-P2-S2 — Create and restore practice
|
|
|
|
- **Outcome:** A student creates and restores a tenant-scoped practice session from valid published content without client-supplied identity or tenant IDs.
|
|
- **Risk:** Medium; session ownership, graph scope, and schema packaging require negative tests.
|
|
- **Blockers:**
|
|
- EDU-P0-S1 safe-content contract.
|
|
- Core-loop schema must be promoted into active module-owned Flyway history.
|
|
- Provider authority and existing dirty session implementation classification.
|
|
- **Verification:**
|
|
- Tenant/user context and Member-principal tests.
|
|
- Restore ownership, cross-tenant denial, PUBLIC-scope tests.
|
|
- PostgreSQL uniqueness, transaction, packaging, and migration execution checks if schema changes are approved.
|
|
|
|
## EDU-P2-S3 — Idempotent answer save
|
|
|
|
- **Outcome:** A student saves one answer idempotently with explicit duplicate/conflicting-payload semantics and no sensitive-field exposure.
|
|
- **Risk:** Medium-to-high due to concurrent writes, stale versions, and answer leakage.
|
|
- **Blockers:**
|
|
- EDU-P2-S2 session state.
|
|
- Existing answer/idempotency schema and option snapshot contract.
|
|
- Entitlement decision for non-public/private content.
|
|
- **Verification:**
|
|
- Same-payload duplicate and conflicting-payload tests.
|
|
- Concurrent PostgreSQL uniqueness/transaction tests.
|
|
- Tenant isolation, stale-version, safe-response, and malformed-snapshot tests.
|
|
|
|
## EDU-P2-S4 — Atomic submit, report, wrong questions, and favorites
|
|
|
|
- **Outcome:** A student atomically claims and submits a session, reads an immutable report, and receives consistent wrong-question/favorite projections.
|
|
- **Risk:** High; current check-then-insert submit idempotency is not sufficient.
|
|
- **Blockers:**
|
|
- EDU-P2-S3 answer state.
|
|
- Atomic submit-key reservation and crash recovery design.
|
|
- Scoring/report immutability and entitlement decisions.
|
|
- **Verification:**
|
|
- ON CONFLICT/atomic claim concurrency tests.
|
|
- Processing-row crash recovery and retry semantics.
|
|
- Immutable report/scoring, duplicate submission, wrong-question/favorite idempotency, sensitive-field, tenant, and unauthorized tests.
|
|
|
|
## EDU-P3-S5 — Tenant content publication and graph integrity
|
|
|
|
- **Outcome:** Tenant administrators author, classify, publish, and safely retire question content with tenant-consistent graph integrity.
|
|
- **Risk:** High due to publication, admin scope, public graph, and student-read consistency.
|
|
- **Blockers:**
|
|
- Core loop verified.
|
|
- Provider and education content model decisions.
|
|
- System RBAC/DataPermission policy.
|
|
- **Verification:**
|
|
- Admin permission/row-scope tests.
|
|
- Composite tenant/scope relationship constraint or equivalent enforcement tests.
|
|
- Publication visibility/provider consistency and safe-projection regression tests.
|
|
|
|
## EDU-P3-S6 — Content imports, exports, assets, and scanning
|
|
|
|
- **Outcome:** Tenant administrators import/export education content with durable business state, leases, retries, duplicate-safe processing, file security, and audit.
|
|
- **Risk:** High operational and security risk.
|
|
- **Blockers:**
|
|
- Publication model.
|
|
- Infra File contract and scanner ownership.
|
|
- Infra Job/MQ durable claim/lease semantics.
|
|
- **Verification:**
|
|
- Preview/execute state machine.
|
|
- Atomic claim/lease/heartbeat/expiry/retry/dead-letter tests.
|
|
- MIME/size/object-key/scan fail-closed tests.
|
|
- Tenant propagation, audit redaction, and partial-failure tests.
|
|
|
|
## EDU-P4-S7 — Classes and education relationships
|
|
|
|
- **Outcome:** Tenant administrators manage classes, education student relationships, invitations, supervision, and education operations with explicit scope.
|
|
- **Risk:** High authorization risk.
|
|
- **Blockers:**
|
|
- Education relationship model.
|
|
- System RBAC/DataPermission scope rules.
|
|
- Member relationship contract and CRM supervision decision.
|
|
- **Verification:**
|
|
- Student/teacher/class permission matrix.
|
|
- Cross-class/cross-tenant negative and duplicate invitation tests.
|
|
- Audit redaction and operation-log tests.
|
|
|
|
## EDU-P4-S8 — Tenant configuration, integrations, and access operations
|
|
|
|
- **Outcome:** Selected tenant appearance, integrations, secrets, activation codes, coupons, and public-bank access capabilities have explicit owners and safe contracts.
|
|
- **Risk:** High because secret, payment configuration, redemption, and public-bank synchronization boundaries differ.
|
|
- **Blockers:**
|
|
- Appearance/domain/integration/secrets/codes ownership decisions.
|
|
- System tenant configuration and secret APIs.
|
|
- Mall/Pay/Member entitlement and code contracts.
|
|
- **Verification:**
|
|
- Secret redaction/rotation and authorization tests.
|
|
- Domain/auth-provider/payment-account configuration tests.
|
|
- Code/coupon redemption idempotency and audit tests.
|
|
- Public-bank grant/sync and cross-tenant tests.
|
|
|
|
## EDU-P5-S9 — Education commercialization binding
|
|
|
|
- **Outcome:** Education products bind to commerce purchases and Member entitlements without duplicated financial ledgers.
|
|
- **Risk:** High financial and authorization risk.
|
|
- **Blockers:**
|
|
- Product binding model.
|
|
- Mall/Pay public APIs.
|
|
- Member entitlement decision and callback/refund semantics.
|
|
- **Verification:**
|
|
- Order/payment/refund callback contracts.
|
|
- Entitlement issuance/revocation/expiry and idempotent fulfillment.
|
|
- Reconciliation, commission/referral, authorization, and audit tests.
|
|
|
|
## EDU-P5-S10 — Extended student and secondary learning waves
|
|
|
|
- **Outcome:** Selected Auth/Profile/extended Learning/scoreline/vocabulary/video/AI/notification/badge/exam capabilities are migrated, replaced, retired, or deferred with traceable decisions.
|
|
- **Risk:** Medium-to-high due to omitted student contracts, media entitlement, and unclear ownership.
|
|
- **Blockers:**
|
|
- Explicit scope for each secondary capability.
|
|
- AI/File/Member/System/Infra contracts and entitlement model.
|
|
- **Verification:**
|
|
- Per-capability endpoint/data/authorization contract tests.
|
|
- Progress/report/vocabulary state tests.
|
|
- Media entitlement, safe export/redaction, tenant isolation, and retirement compatibility tests.
|
|
|
|
## EDU-P6-S11 — Operational independence and legacy exit
|
|
|
|
- **Outcome:** Background and platform operations run independently of NestJS with documented retries, scanning, audit, notifications, observability, and deployment evidence.
|
|
- **Risk:** High deployment and reliability risk.
|
|
- **Blockers:**
|
|
- All owner and contract decisions.
|
|
- Operational deployment, scanner, observability, and legacy exit plan.
|
|
- **Verification:**
|
|
- Worker/job deployment smoke tests.
|
|
- At-least-once duplicate/dead-letter and scanner health/security tests.
|
|
- PostgreSQL migration execution evidence.
|
|
- Runbook and documentation consistency review.
|