Skip to main content

Slice Risk Register

:::info Source Sourced from docs/roadmap/slice-risk-register.md in the documentation repo. :::

Execution-layer companion to ROADMAP.md and 14 Risks & Trade-offs.

Risks framed per slice (S0–S6) so risk discussions are scheduled alongside work that might trigger them. Each row: severity (S1 critical · S2 high · S3 medium · S4 low), impact, mitigation, owner, dependencies.

S0 — Platform Foundation

IDRiskSevImpactMitigationOwnerDependency
S0-R1Tenant isolation regressionS1Cross-tenant data leak; contract terminationsTwo-tenant CI suite; mandatory code review on any RLS policy change; pen-test; RLS bypass testsPlatform + SecurityPostgres RLS framework; JWT + RequestContext
S0-R2Event envelope driftS2Services fall out of sync; Pact breakage; refactor avalancheEnvelope frozen + schema registry CI gate; ADR for any changePlatformSchema registry
S0-R3AI gateway port contract churnS219 services refactorAIClient port frozen; version rule additive-only; adapter abstractionAI ServicesAI adapter tests
S0-R4Sync protocol churnS2Every client rebuilds; offline bundles invalidated/sync/v1/ frozen; additive-onlySync + PlatformSync protocol ADR
S0-R5KMS mis-configurationS1Loss of data confidentiality; DR failureKMS key hierarchy + rotation design reviewed by Security; DR drillSecurity + PlatformKMS vendor selection
S0-R6OpenTelemetry overheadS3Unexpected latency or costSampling; async exporters; dashboards for OTel healthSRE
S0-R7Two-tenant test suite gapsS1Silent leaks slip past CIMatrix-test every endpoint; property-based testsPlatform + SecurityCI infra
S0-R8Over-building the foundationS2Time-to-M1 slipsStrict M0 scope doc; weekly backlog reviewPM + Platform lead

S1 — Minimal Learner (M1)

IDRiskSevImpactMitigationOwnerDependency
S1-R1Offline bundle tamper/device-binding bugS1Content piracy; license bypassAES-256-GCM per-device derivation; JWS signing; tamper CI fixtures; bundle chaos testsContent + SecurityKMS per tenant; device cert
S1-R2AI tutor hallucination at learner surfaceS1Wrong answers in compliance training; regulatory exposureRAG over lesson context; refusal UX; citation of cited blocks; quarterly accuracy evalAI + LearningPrompt registry; eval harness
S1-R3Local model quality gapS2Offline UX feels degraded"Local model" badge; cloud-refresh CTA; quality eval per releaseAI + Mobile PlatformLocal-inference SDK
S1-R4PlayPackage schema late freezeS2Player + Content-Packaging divergeFreeze before M1 sprint 1; shared TS typesContent-Packaging + LearnerBlock schema
S1-R5License envelope expiry UX ambiguityS2Learners blocked without explanationClear countdown UX; proactive refresh on syncLearner + DesignSync service
S1-R6Statement outbox overflow on long offline periodsS2Lost statementsChunked push; client-side caps; backpressure UXSync + ProgressIndexedDB quotas
S1-R7Multi-device cursor resolution bugS2Learner confused about progressmax(cursor) reconciliation + tests; audit each reconciliationLearner + SyncVector clock
S1-R8Accessibility regressions on playerS2WCAG 2.2 AA failureaxe in CI; manual NVDA + VoiceOver per release; reduced-motion toggleDesign + Learner FE
S1-R9Capacitor ↔ web parity gapsS3Bugs appear only on mobileShared E2E fixtures; device farm testsMobile Platform + QADevice farm
S1-R10Design partners insufficient diversityS3Missed feedback from regulated/remote usersCurate partner cohort (regulated, field, multilingual)PM + Sales

S2 — Authoring MVP + AI Co-Author MVP (M2 first half)

IDRiskSevImpactMitigationOwnerDependency
S2-R1Publish saga half-failuresS1Orphan CourseVersions, broken catalogExplicit compensations; chaos tests at every step; admin queue; saga state machine testsPlatform + Authoring + ContentSaga infra
S2-R2Block registry rushedS1Block kind shape churn + reworkBlock schema RFC + freeze at M2 start; new kinds additive onlyAuthoring + ArchitectureBlock schema ADR
S2-R3AI co-author accept rate lowS2Low adoption, wasted AI spendPrompt regression gate at 50 % accept; user-research cadenceAI + AuthoringEval harness
S2-R4Provenance UI complexityS3Admins ignore AI transparencyProgressive disclosure in UI; badge always visibleDesign + Authoring FE
S2-R5Media pipeline bottleneck on transcodeS2Slow author feedback loopWorker pool; backpressure UX; inline low-res previewMedia
S2-R6Customer content shape surprisesS3Real content breaks block validatorsPartner beta with real content before freezeAuthoring + PM
S2-R7Publish saga retries exhaust AI budgetS3Unexpected AI cost spikeIdempotent AI calls; cache by prompt-hash; retry capsAI + Authoring

S3 — Marketplace MVP (M2 second half)

IDRiskSevImpactMitigationOwnerDependency
S3-R1Payment compliance gapsS1PCI incident; processor terminationTokenized cards only; PCI scope minimized; processor-abstract ACLCommerce + SecurityProcessor sandbox
S3-R2Refund edge cases leak seatsS2Provider disputes; partial refunds wrongRefund policy DSL + unit-tested matrix; refund-after-seat-consumed ruleCommerce + Legal
S3-R3SCORM 1.2 conformance regressionS23rd-party LMS rejects zipsSCORM Cloud in CI every build; fixture coursesContent-PackagingSCORM Cloud account
S3-R4Webhook replay storms from customersS3DLQ + alert fatigueBackoff + DLQ + dashboards; per-subscription limitsComms + SRE
S3-R5Marketplace low-quality listings at launchS2Brand damageAI moderation + human review; provider onboarding standardsCommerce + AIModeration pipeline
S3-R6Purchase saga split-brain with licensingS1Payment without license or license without paymentIdempotent saga + compensations + reconciliation jobCommerce + Platform
S3-R7Public certificate verify abused for scrapingS3Data harvestRate limit + bot mitigation + verification-token TTL schemeCertification + Security

S4 — Compliance + Enterprise (M3)

IDRiskSevImpactMitigationOwnerDependency
S4-R1RRULE + timezone correctnessS2Wrong due dates; compliance failures1 000-fixture suite incl. DST + leap; TZ matrix testsEnterpriseRRULE engine
S4-R2SAML edge cases per IdPS2Enterprise deals stallTest Okta, Azure AD, Google, custom ADFS, Auth0Enterprise + PlatformIdP test accounts
S4-R3ABAC policy complexity breeds mis-grantsS1Data leak within tenantPolicy linter; sample-data tests; UI shows plain-language policyPlatformABAC DSL
S4-R4AI grading fairnessS1Discrimination claimsBias eval; human override; EU AI Act high-risk docs; external auditAI + ComplianceEval corpus
S4-R5PDF→course quality variableS2Authors reject AI outputConfidence thresholds; chunk-level accept/reject; fallback to outline-onlyAI + Authoring
S4-R6Recurrence storm (many tenants activate on same day)S2Notification burst + queue overloadJitter materialization; batch send; backpressureEnterprise + Comms
S4-R7GDPR erasure saga driftS1Erasure incomplete; regulator riskEvery service declares participation; CI gate; saga replay testsPlatform + ComplianceGDPR saga contract
S4-R8SCORM 2004 + xAPI conformance missesS2Regulated market rejectionsADL suite in CI; cmi5 profile testsContent-PackagingADL LRS
S4-R9Enterprise procurement delaysS2Revenue slipsSOC 2 Type I + DPA + BAA templates ready; reference customersEnterprise + LegalSOC 2 auditor

S5 — Full Authoring + Offline Authoring (M4)

IDRiskSevImpactMitigationOwnerDependency
S5-R1Offline authoring conflict UXS1Data loss perceptionPre-merge backup; side-by-side diff; AI merge suggestion; 30-day backup retentionAuthoring + SyncConflict UI
S5-R2Yjs doc corruptionS2Collab session lostPeriodic snapshots; replay from event log; conflict repair toolingAuthoringYjs persistence
S5-R3Live-collab latency across regionsS2UX feels laggyRegional WS endpoints; presence throttle; awareness compressionAuthoring + SRE
S5-R4AI image/TTS content-safety + copyrightS2Legal exposureContent-safety pipeline; provenance on every asset; copyright-risk classifierMedia + AI + Legal
S5-R5LTI 1.3 interop quirksS2Embedding deals stallLTI conformance tests; partner sandboxEnterprise + TenantLTI tooling
S5-R6Block taxonomy bloatS2Editor UX complexityGovernance board; usage telemetry; quarterly pruneAuthoring + Design
S5-R7Hybrid search ranker qualityS2Low relevanceEval with user-judged pairs; A/B ranker rolloutData/AI + Search
S5-R8AI translation errors on regulated terminologyS2Legal riskPer-tenant glossaries; reviewer required; legal-language flagAI + AuthoringGlossary tooling

S6 — Scale + Advanced Insight + Mobile (M5)

IDRiskSevImpactMitigationOwnerDependency
S6-R1Multi-region data residency migration bugsS1Data loss or cross-region leakageRehearsals on production-size fixture; checksum verification; rollback path; saga testsPlatform + SRE + all servicesResidency saga
S6-R2HIPAA provider allowlist enforcementS1BAA non-complianceTenant-tagged routing; CI gate on provider list; audit exportAI Services + ComplianceBAA contracts
S6-R3Mobile native regressions from platform updatesS2App-store rejectionDevice farm; beta channel; staged rolloutMobile Platform + QADevice farm
S6-R4Marketplace abuse at scaleS2Brand damage; fraud lossAI moderation v2; provider deposits; fraud-signal monitoringCommerce + Security + AI
S6-R5White-label CSP scoping bugsS2XSS across tenantsPer-tenant CSP + nonce; isolated subdomain + cookie scopingPlatform + Security
S6-R6Developer SDK breaking-change temptationsS2Integrator churnSemver strictness; deprecation policy; communication channelsDevEx + PMSDK governance
S6-R7At-risk prediction model biasS1Unfair interventionsQuarterly bias eval; feature exclusion list; human-only override; opt-outData/AI + ComplianceEval corpus
S6-R8ISO 27001 certification scope mismatchS2Audit failControl mapping exercise early; internal audit passCompliance + SREAuditor

Slice-Independent / Cross-Cutting Risks

IDRiskSevImpactMitigationOwner
X-R1AI cost runawayS1Surprise billsPer-tenant budgets + soft-degrade + hard-stop + alertsAI Services + Finance
X-R2Over-eager AI defaults reduce trustS2Users distrust productDefault OFF per tenant; per-feature opt-in; transparent provenanceAI + Design
X-R3Schema drift across servicesS2Pact breakageSchema registry; CI gate; weekly producer reviewPlatform
X-R4Solo on-call burnoutS2Incident response quality dropsRotation; buddy system; post-incident reviews weeklySRE
X-R5Regional compliance surprisesS2Launch blockersLegal-review per geo before launchLegal + PM
X-R6Pilot feedback overwrites roadmapS3Scope creepPM triage; feedback lands in backlog with slice assignmentPM

Governance

  • Weekly risk review: each team owner presents new, changed, or closed risks.
  • Quarterly architecture risk review: top 10 cross-cutting risks reviewed by CTO + architecture.
  • Every S1 risk has a named owner, a due date for mitigation, and a verification plan.
  • Acceptance criteria for S1/S2 risks to be "closed" includes: mitigation shipped + metric(s) monitored + post-mitigation verification test documented.