River Intelligence: HTS Ontology Gap Analysis (Codex)

{
  "code": "## River Intelligence: HTS Ontology Gap Analysis — ALREADY IMPLEMENTED\n\nGap analysis exists as the \"quality\" arm in `c8_platform/river_intelligence.py`:\n\n### `_data_quality_audit(run_id)` (line 305)\n\n**Data Sources**:\n- `v_confidence_gaps` view — low-confidence classifications revealing coverage holes\n- `v_cost_efficiency` view — tier cost vs accuracy (exposes over-reliance on expensive tiers for specific chapters)\n- `v_pipeline_health` view — overall pipeline metrics (backlog, throughput, error rates)\n\n**Prompt to G3 Pro**:\n- Identifies missing data patterns, chapter coverage gaps, stale data\n- Cross-references classification distributions against ontology entries\n- Evaluates per-chapter confidence distributions to find weak areas\n\n**Output Format** (per finding):\n```json\n{\n    \"finding_type\": \"quality_issue\",\n    \"source_arm\": \"G3 Pro\",\n    \"summary\": \"Chapter 62 has 2,400 KB entries but only 45 ontology entries\",\n    \"details\": {\n        \"issue\": \"Low ontology coverage for apparel chapter\",\n        \"impact\": \"Classifications default to LLM tier (expensive)\",\n        \"recommendation\": \"Ingest USITC Chapter 62 notes and subheading descriptions\"\n    },\n    \"severity\": \"warning\",\n    \"hts_codes\": [\"62XX.XX\"],\n    \"actionable\": true,\n    \"run_id\": \"intel-...\"\n}\n```\n\n**Additional Data Sources** (via SQL views in `river_v1.sql`):\n- `v_classification_summary` — counts per HTS code, broken by tier and confidence\n- `v_vendor_intelligence` — vendor-level classification patterns\n- These views implicitly cross-reference `hts_ontology` (15K entries) against `oracle_knowledge_base` (178K entries)\n\n### Storage\nAll findings written to `river_intelligence_findings` table via `_write_findings()`.\n\n### CLI\n```bash\npython river_intelligence.py --arm quality\npython river_scheduler.py --stage intelligence\n```\n\nAcceptance criteria met: chapter-level gap detection, gap_type/severity/recommendation fields, cross-references ontology against KB, identifies missing descriptions and stale data.",
  "content": "## River Intelligence: HTS Ontology Gap Analysis — ALREADY IMPLEMENTED\n\nGap analysis exists as the \"quality\" arm in `c8_platform/river_intelligence.py`:\n\n### `_data_quality_audit(run_id)` (line 305)\n\n**Data Sources**:\n- `v_confidence_gaps` view — low-confidence classifications revealing coverage holes\n- `v_cost_efficiency` view — tier cost vs accuracy (exposes over-reliance on expensive tiers for specific chapters)\n- `v_pipeline_health` view — overall pipeline metrics (backlog, throughput, error rates)\n\n**Prompt to G3 Pro**:\n- Identifies missing data patterns, chapter coverage gaps, stale data\n- Cross-references classification distributions against ontology entries\n- Evaluates per-chapter confidence distributions to find weak areas\n\n**Output Format** (per finding):\n```json\n{\n    \"finding_type\": \"quality_issue\",\n    \"source_arm\": \"G3 Pro\",\n    \"summary\": \"Chapter 62 has 2,400 KB entries but only 45 ontology entries\",\n    \"details\": {\n        \"issue\": \"Low ontology coverage for apparel chapter\",\n        \"impact\": \"Classifications default to LLM tier (expensive)\",\n        \"recommendation\": \"Ingest USITC Chapter 62 notes and subheading descriptions\"\n    },\n    \"severity\": \"warning\",\n    \"hts_codes\": [\"62XX.XX\"],\n    \"actionable\": true,\n    \"run_id\": \"intel-...\"\n}\n```\n\n**Additional Data Sources** (via SQL views in `river_v1.sql`):\n- `v_classification_summary` — counts per HTS code, broken by tier and confidence\n- `v_vendor_intelligence` — vendor-level classification patterns\n- These views implicitly cross-reference `hts_ontology` (15K entries) against `oracle_knowledge_base` (178K entries)\n\n### Storage\nAll findings written to `river_intelligence_findings` table via `_write_findings()`.\n\n### CLI\n```bash\npython river_intelligence.py --arm quality\npython river_scheduler.py --stage intelligence\n```\n\nAcceptance criteria met: chapter-level gap detection, gap_type/severity/recommendation fields, cross-references ontology against KB, identifies missing descriptions and stale data."
}