{"slug":"documentation/MISSING_SERVICES_IMPLEMENTATION_GUIDE","title":"Missing Canonical Services Implementation Guide","section":"documentation","sectionLabel":"Documentation","status":"unclassified","type":"document","path":"documentation/MISSING_SERVICES_IMPLEMENTATION_GUIDE.mdx","bytes":4511,"description":null,"updated":null,"source":"integratewise/Integratewise-docs-wiki","content":"\n> **Status: PROPOSED**\n>\n> **Approved: NO**\n>\n> **Execution Authority: NONE**\n>\n> **Evidence Authority: `POST_STOP_FORENSIC_RECONCILIATION.md`**\n\n**Original Status Claim**: Services identified as missing during canonical restoration audit — **UNAPPROVED PROPOSAL**  \n**Original Impact Claim**: Customer Zero activation is blocked on these implementations — **NOT VERIFIED AS AN EXECUTION PREREQUISITE**\n\n---\n\n## Missing Service 1: spine-writer\n\n**Purpose**: Canonical write path for the Operational Spine  \n**Current Workaround**: `integratewise-pipeline` may be handling writes (legacy role)  \n**Required Inputs**:\n\n- Entity mutations from workflows, actions, decisions\n- Spine schema (entity types, relationships)\n- Neon Fortress connection for canonical writes; D1/KV only for bounded\n  operational state\n- Continuity event publishing for audit\n\n**Required Outputs**:\n\n- Successful write confirmation (entity_id, version, timestamp)\n- Continuity record created for recovery\n- Timeline event published\n- Cache invalidation for dependents\n\n**Implementation Steps**:\n\n1. Create `services/spine-writer/src/index.ts`\n2. Implement entity write handler (POST /entities)\n3. Connect the governed writer to Neon Fortress and emit bounded D1/KV\n   invalidation/rehydration events\n4. Publish continuity events\n5. Deploy as `integratewise-spine-writer`\n\n---\n\n## Missing Service 2: governance\n\n**Purpose**: Policy evaluation, approval chains, governance decisions  \n**Current Workaround**: `integratewise-govern` (wrong name) may have partial logic  \n**Required Inputs**:\n\n- Capability proposal or action request\n- Policy configuration for tenant\n- User/role context\n- Historical decision records\n\n**Required Outputs**:\n\n- Governance decision (approved, rejected, requires_approval)\n- Approval record created\n- Policy audit trail\n\n**Implementation Steps**:\n\n1. Create `services/governance/src/index.ts` (note: correct spelling)\n2. Load tenant policies from policy store\n3. Evaluate against capability request\n4. Route to approval chain if needed\n5. Deploy as `integratewise-governance`\n\n**Note**: Rename or deprecate `integratewise-govern` to avoid naming conflict.\n\n---\n\n## Missing Service 3: schema-synthesis\n\n**Purpose**: AI-driven schema inference and generation  \n**Current Workaround**: None (critical gap)  \n**Required Inputs**:\n\n- Raw data samples (JSON, CSV, unstructured)\n- Existing schema patterns (for similarity detection)\n- User hints or type guidance\n\n**Required Outputs**:\n\n- Inferred schema (entity types, field mappings, relationships)\n- Confidence scores\n- Schema validation rules\n\n**Implementation Steps**:\n\n1. Create `services/schema-synthesis/src/index.ts`\n2. Integrate HuggingFace inference for schema generation\n3. Store inferred schemas in store service\n4. Cache common patterns\n5. Deploy as `integratewise-schema-synthesis`\n\n---\n\n## Missing Service 4: huggingface-inference\n\n**Purpose**: Inference endpoint for AI model calls  \n**Current Workaround**: Calls may be routed directly to HuggingFace (no abstraction)  \n**Required Inputs**:\n\n- Model ID (from HuggingFace)\n- Input prompt or data\n- Parameters (temperature, max_tokens, etc.)\n\n**Required Outputs**:\n\n- Model output (text, embeddings, predictions)\n- Inference latency\n- Token usage for billing\n\n**Implementation Steps**:\n\n1. Create `services/huggingface-inference/src/index.ts`\n2. Implement model inference handler\n3. Support multiple model types (text generation, embeddings, classification)\n4. Add caching layer for common queries\n5. Deploy as `integratewise-huggingface-inference`\n\n---\n\n## Timeline to Full Canonical Readiness\n\n| Service               | Effort | Priority | Dependency                    |\n| --------------------- | ------ | -------- | ----------------------------- |\n| spine-writer          | Medium | Critical | Needed for entity mutations   |\n| governance            | Medium | High     | Needed for approval chains    |\n| schema-synthesis      | High   | Medium   | Needed for schema AI features |\n| huggingface-inference | Low    | High     | Needed for all inference      |\n\n**Estimated total**: 3–5 days for all four services\n\n---\n\n## Customer Zero Activation Can Begin With:\n\n- 12 deployed canonical services (currently live)\n- Basic workflow execution (without governance approval chain)\n- Data normalization and querying (without schema synthesis)\n- Inference calls routed directly or through adapter\n\n**Full feature parity requires all 18 services.**\n"}