smoke: 3 docs + vectors

chunks.jsonl

@@ -0,0 +1,5 @@
+{"chunk_index": 0, "context_prefix": "This document outlines the key terms of umbrella policy PES-2024-00312, specifying $25 million in excess coverage limits over three underlying insurance lines (CGL, auto liability, and employers liability), clarifying that the umbrella only fills gaps in existing coverages without expanding into excluded areas like pollution or cyber liability, and establishing a 30-day claims notification requirement with a 60-day extension for criminal investigations.", "created_at": "2026-06-24T07:55:15.443972+00:00", "document_id": "494dcc8b-6f47-4d12-a788-bccf12691b80", "id": "4a906dff-a258-4de9-beb4-03bf9dd73a30", "metadata": {}, "page_end": 1, "page_start": 1, "section_title": "Umbrella Policy Terms — PES-2024-00312", "text": "# Umbrella Policy Terms — PES-2024-00312\n\nThe umbrella policy provides $25,000,000 of excess limits over the\nfollowing underlying lines:\n\n* CGL — first $2,000,000 aggregate\n* Auto liability — first $1,000,000 per accident\n* Employers liability — first $1,000,000 per occurrence\n\nThe umbrella drops down to fill gaps only for coverages the underlying\npolicies provide. It does not broaden into pollution liability or\ncyber, even if the underlying policies were silent.\n\nNotification window for claims is 30 days from the insured’s discovery,\nextended to 60 days for ongoing investigations under criminal subpoena.", "token_count": 163, "workspace_id": "harvey-extract"}
+{"chunk_index": 0, "context_prefix": "This is the 2024 property insurance schedule for Cascade's four-location portfolio, detailing total insured values ranging from $7.8M to $18.4M per location, with deductibles between $25K–$100K. All buildings use replacement cost valuation with 90% coinsurance requirements. Flood coverage is capped at $5M per occurrence across the portfolio, and earthquake coverage is excluded at the Boise Datacenter due to its seismic zone classification.", "created_at": "2026-06-24T07:55:08.985578+00:00", "document_id": "452bcf31-97a8-432e-8ad6-c4edbdeba2ed", "id": "660e2bbb-d61a-4f54-b9b6-150585e7d60d", "metadata": {}, "page_end": 1, "page_start": 1, "section_title": "Property Schedule of Insurance — 2024", "text": "# Property Schedule of Insurance — 2024\n\nThe schedule covers four locations across the Cascade portfolio. Each\nbuilding is valued at replacement cost with a 90% coinsurance clause.\n\n| Location | TIV (USD) | Deductible |\n|---|---|---|\n| Seattle HQ | 18,400,000 | 50,000 |\n| Tacoma Warehouse | 9,200,000 | 25,000 |\n| Portland Branch | 7,800,000 | 25,000 |\n| Boise Datacenter | 12,100,000 | 100,000 |\n\nFlood is sub-limited to $5,000,000 per occurrence and earthquake is\nexcluded for the Boise location due to seismic zone classification.", "token_count": 187, "workspace_id": "harvey-extract"}
+{"chunk_index": 2, "context_prefix": "This CGL policy (GNM-2024-88714) excludes pollution liability, professional services errors and omissions, and cyber/data breach costs, which are covered under separate specialized policies. The policy provides $2M aggregate and $1M per-occurrence limits with a $25K deductible per occurrence.", "created_at": "2026-06-24T07:55:01.077586+00:00", "document_id": "942c3ac5-467f-4b9f-92ca-cbbb2785c634", "id": "7cde98aa-d6f2-4d42-a8b5-b033e8e7cd5b", "metadata": {}, "page_end": 1, "page_start": 1, "section_title": "Exclusions", "text": "## Exclusions\n\n* Pollution liability (handled by separate environmental policy).\n* Professional services E&O.\n* Cyber events and data breach response costs.", "token_count": 42, "workspace_id": "harvey-extract"}
+{"chunk_index": 0, "context_prefix": "Commercial General Liability policy GNM-2024-88714 provides coverage for bodily injury and property damage from premises operations, with a $2,000,000 aggregate limit per policy period and $1,000,000 per occurrence limit. The policy includes a $25,000 deductible per occurrence and excludes pollution liability, professional services errors and omissions, and cyber-related claims.", "created_at": "2026-06-24T07:55:01.077567+00:00", "document_id": "942c3ac5-467f-4b9f-92ca-cbbb2785c634", "id": "923adc14-91a9-4120-9baa-5d2c48cf7c9e", "metadata": {}, "page_end": 1, "page_start": 1, "section_title": "CGL Policy Summary — GNM-2024-88714", "text": "# CGL Policy Summary — GNM-2024-88714\n\nThis Commercial General Liability policy covers bodily injury and\nproperty damage claims arising from premises operations. The aggregate\nlimit is $2,000,000 per policy period and $1,000,000 per occurrence.", "token_count": 69, "workspace_id": "harvey-extract"}
+{"chunk_index": 1, "context_prefix": "This CGL policy requires the insured to pay a $25,000 deductible per occurrence before coverage activates. The deductible reduces the $1,000,000 per-occurrence limit but does not affect the $2,000,000 aggregate limit, meaning multiple claims can accumulate toward the aggregate without additional deductible reductions.", "created_at": "2026-06-24T07:55:01.077581+00:00", "document_id": "942c3ac5-467f-4b9f-92ca-cbbb2785c634", "id": "e58ec86a-926c-4fe7-a191-07654810ff0f", "metadata": {}, "page_end": 1, "page_start": 1, "section_title": "Deductible", "text": "## Deductible\n\nThe insured pays a $25,000 deductible per occurrence before coverage\napplies. The deductible erodes the per-occurrence limit but not the\naggregate.", "token_count": 50, "workspace_id": "harvey-extract"}

documents.jsonl

@@ -0,0 +1,3 @@
+{"created_at": "2026-06-24T07:55:08.984793+00:00", "error": null, "filename": "property-schedule.md", "id": "452bcf31-97a8-432e-8ad6-c4edbdeba2ed", "metadata": {}, "mime_type": "text/markdown", "page_count": 1, "parsed_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/452bcf31-97a8-432e-8ad6-c4edbdeba2ed/parsed.json", "sha256": "03ff998873db8e74d06ff1af0a7e755707e1422ffbb52106a011103ad57971f3", "size_bytes": 537, "stage": "ready", "storage_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/452bcf31-97a8-432e-8ad6-c4edbdeba2ed/raw/property-schedule.md", "summary": "The 2024 Property Schedule of Insurance covers four Cascade portfolio locations: Seattle HQ ($18.4M), Tacoma Warehouse ($9.2M), Portland Branch ($7.8M), and Boise Datacenter ($12.1M), with total insured value of $47.5M. All buildings are valued at replacement cost with a 90% coinsurance clause and varying deductibles ($25K–$100K). Flood coverage is sub-limited to $5M per occurrence, and earthquake coverage is excluded for the Boise location due to its seismic zone classification.", "title": "property-schedule.md", "topics": ["Property Schedule 2024", "Cascade portfolio", "Total Insured Value $47.5M", "Replacement cost valuation", "90% coinsurance clause", "Seattle HQ $18.4M", "Tacoma Warehouse $9.2M", "Portland Branch $7.8M", "Boise Datacenter $12.1M", "Flood sub-limit $5M", "Earthquake exclusion Boise", "Seismic zone classification", "Deductibles $25K-$100K"], "updated_at": "2026-06-24T07:55:15.427699+00:00", "workspace_id": "harvey-extract"}
+{"created_at": "2026-06-24T07:55:15.443303+00:00", "error": null, "filename": "umbrella-terms.md", "id": "494dcc8b-6f47-4d12-a788-bccf12691b80", "metadata": {}, "mime_type": "text/markdown", "page_count": 1, "parsed_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/494dcc8b-6f47-4d12-a788-bccf12691b80/parsed.json", "sha256": "a468d44ddfc1b176b86cce9d800e60db140dac165d7ccaf4f29423bd798e801d", "size_bytes": 623, "stage": "ready", "storage_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/494dcc8b-6f47-4d12-a788-bccf12691b80/raw/umbrella-terms.md", "summary": "Umbrella policy PES-2024-00312 provides $25,000,000 in excess limits over underlying commercial general liability ($2,000,000 aggregate), auto liability ($1,000,000 per accident), and employers liability ($1,000,000 per occurrence) coverages. The umbrella only fills gaps in coverages the underlying policies provide and does not expand into pollution liability or cyber coverage. Claims must be reported within 30 days of discovery, with an extension to 60 days if investigations are ongoing under criminal subpoena.", "title": "umbrella-terms.md", "topics": ["umbrella policy", "excess limits", "commercial general liability", "auto liability", "employers liability", "coverage gaps", "pollution liability exclusion", "cyber exclusion", "claims notification", "30-day reporting window", "60-day extension", "criminal subpoena", "PES-2024-00312"], "updated_at": "2026-06-24T07:55:20.756890+00:00", "workspace_id": "harvey-extract"}
+{"created_at": "2026-06-24T07:55:01.011314+00:00", "error": null, "filename": "cgl-summary.md", "id": "942c3ac5-467f-4b9f-92ca-cbbb2785c634", "metadata": {}, "mime_type": "text/markdown", "page_count": 1, "parsed_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/942c3ac5-467f-4b9f-92ca-cbbb2785c634/parsed.json", "sha256": "b9e742f9f62cfbb590f5a0bf30db3fd53977056dac65b2542e22680d7abc527f", "size_bytes": 569, "stage": "ready", "storage_uri": "file:///private/tmp/sagalog-test/A/data/artifacts/harvey-extract/942c3ac5-467f-4b9f-92ca-cbbb2785c634/raw/cgl-summary.md", "summary": "This Commercial General Liability policy (GNM-2024-88714) provides coverage for bodily injury and property damage claims from premises operations with a $2,000,000 aggregate limit and $1,000,000 per-occurrence limit. The insured is responsible for a $25,000 deductible per occurrence, which reduces the per-occurrence limit but not the aggregate. Key exclusions include pollution liability (covered separately), professional services errors and omissions, and cyber events or data breach response costs.", "title": "cgl-summary.md", "topics": ["Commercial General Liability", "CGL Policy", "Bodily Injury", "Property Damage", "Premises Operations", "$2M Aggregate Limit", "$1M Per-Occurrence Limit", "$25K Deductible", "Pollution Exclusion", "Professional Services E&O Exclusion", "Cyber Exclusion", "Data Breach Exclusion", "GNM-2024-88714"], "updated_at": "2026-06-24T07:55:08.969312+00:00", "workspace_id": "harvey-extract"}

parsed/452bcf31-97a8-432e-8ad6-c4edbdeba2ed/doc.md

@@ -0,0 +1,17 @@
+## Property Schedule of Insurance — 2024
+
+# Property Schedule of Insurance — 2024
+
+The schedule covers four locations across the Cascade portfolio. Each
+building is valued at replacement cost with a 90% coinsurance clause.
+
+| Location | TIV (USD) | Deductible |
+|---|---|---|
+| Seattle HQ | 18,400,000 | 50,000 |
+| Tacoma Warehouse | 9,200,000 | 25,000 |
+| Portland Branch | 7,800,000 | 25,000 |
+| Boise Datacenter | 12,100,000 | 100,000 |
+
+Flood is sub-limited to $5,000,000 per occurrence and earthquake is
+excluded for the Boise location due to seismic zone classification.
+

parsed/452bcf31-97a8-432e-8ad6-c4edbdeba2ed/index.jsonl

@@ -0,0 +1 @@
+{"chunk_id": "660e2bbb-d61a-4f54-b9b6-150585e7d60d", "chunk_index": 0, "line_end": 16, "line_start": 1, "one_liner": "This is the 2024 property insurance schedule for Cascade's four-location portfolio, detailing total insured values ranging from $7.8M to $18.4M per location, with deductibles between $25K–$100K. All buildings use replacement cost valuation with 90% coinsurance requirements. Flood coverage is capped at $5M per occurrence across the portfolio, and earthquake coverage is excluded at the Boise Datacenter due to its seismic zone classification."}

parsed/452bcf31-97a8-432e-8ad6-c4edbdeba2ed/parsed.json

@@ -0,0 +1 @@
+{"mime": "text/markdown", "page_count": 1, "parser": "markdown", "raw": null, "sections": [{"bbox": [{"height": 14, "left": 0, "page": 1, "text": null, "top": 1, "type": "MarkdownHeading", "width": 0}], "metadata": {}, "page_end": 1, "page_start": 1, "section_type": null, "text": "# Property Schedule of Insurance — 2024\n\nThe schedule covers four locations across the Cascade portfolio. Each\nbuilding is valued at replacement cost with a 90% coinsurance clause.\n\n| Location | TIV (USD) | Deductible |\n|---|---|---|\n| Seattle HQ | 18,400,000 | 50,000 |\n| Tacoma Warehouse | 9,200,000 | 25,000 |\n| Portland Branch | 7,800,000 | 25,000 |\n| Boise Datacenter | 12,100,000 | 100,000 |\n\nFlood is sub-limited to $5,000,000 per occurrence and earthquake is\nexcluded for the Boise location due to seismic zone classification.", "title": "Property Schedule of Insurance — 2024"}]}

parsed/494dcc8b-6f47-4d12-a788-bccf12691b80/doc.md

@@ -0,0 +1,18 @@
+## Umbrella Policy Terms — PES-2024-00312
+
+# Umbrella Policy Terms — PES-2024-00312
+
+The umbrella policy provides $25,000,000 of excess limits over the
+following underlying lines:
+
+* CGL — first $2,000,000 aggregate
+* Auto liability — first $1,000,000 per accident
+* Employers liability — first $1,000,000 per occurrence
+
+The umbrella drops down to fill gaps only for coverages the underlying
+policies provide. It does not broaden into pollution liability or
+cyber, even if the underlying policies were silent.
+
+Notification window for claims is 30 days from the insured’s discovery,
+extended to 60 days for ongoing investigations under criminal subpoena.
+

parsed/494dcc8b-6f47-4d12-a788-bccf12691b80/index.jsonl

@@ -0,0 +1 @@
+{"chunk_id": "4a906dff-a258-4de9-beb4-03bf9dd73a30", "chunk_index": 0, "line_end": 17, "line_start": 1, "one_liner": "This document outlines the key terms of umbrella policy PES-2024-00312, specifying $25 million in excess coverage limits over three underlying insurance lines (CGL, auto liability, and employers liability), clarifying that the umbrella only fills gaps in existing coverages without expanding into excluded areas like pollution or cyber liability, and establishing a 30-day claims notification requirement with a 60-day extension for criminal investigations."}

parsed/494dcc8b-6f47-4d12-a788-bccf12691b80/parsed.json

@@ -0,0 +1 @@
+{"mime": "text/markdown", "page_count": 1, "parser": "markdown", "raw": null, "sections": [{"bbox": [{"height": 15, "left": 0, "page": 1, "text": null, "top": 1, "type": "MarkdownHeading", "width": 0}], "metadata": {}, "page_end": 1, "page_start": 1, "section_type": null, "text": "# Umbrella Policy Terms — PES-2024-00312\n\nThe umbrella policy provides $25,000,000 of excess limits over the\nfollowing underlying lines:\n\n* CGL — first $2,000,000 aggregate\n* Auto liability — first $1,000,000 per accident\n* Employers liability — first $1,000,000 per occurrence\n\nThe umbrella drops down to fill gaps only for coverages the underlying\npolicies provide. It does not broaden into pollution liability or\ncyber, even if the underlying policies were silent.\n\nNotification window for claims is 30 days from the insured’s discovery,\nextended to 60 days for ongoing investigations under criminal subpoena.", "title": "Umbrella Policy Terms — PES-2024-00312"}]}

parsed/942c3ac5-467f-4b9f-92ca-cbbb2785c634/doc.md

@@ -0,0 +1,24 @@
+## CGL Policy Summary — GNM-2024-88714
+
+# CGL Policy Summary — GNM-2024-88714
+
+This Commercial General Liability policy covers bodily injury and
+property damage claims arising from premises operations. The aggregate
+limit is $2,000,000 per policy period and $1,000,000 per occurrence.
+
+## Deductible
+
+## Deductible
+
+The insured pays a $25,000 deductible per occurrence before coverage
+applies. The deductible erodes the per-occurrence limit but not the
+aggregate.
+
+## Exclusions
+
+## Exclusions
+
+* Pollution liability (handled by separate environmental policy).
+* Professional services E&O.
+* Cyber events and data breach response costs.
+

parsed/942c3ac5-467f-4b9f-92ca-cbbb2785c634/index.jsonl

@@ -0,0 +1,3 @@
+{"chunk_id": "923adc14-91a9-4120-9baa-5d2c48cf7c9e", "chunk_index": 0, "line_end": 7, "line_start": 1, "one_liner": "Commercial General Liability policy GNM-2024-88714 provides coverage for bodily injury and property damage from premises operations, with a $2,000,000 aggregate limit per policy period and $1,000,000 per occurrence limit. The policy includes a $25,000 deductible per occurrence and excludes pollution liability, professional services errors and omissions, and cyber-related claims."}
+{"chunk_id": "e58ec86a-926c-4fe7-a191-07654810ff0f", "chunk_index": 1, "line_end": 15, "line_start": 9, "one_liner": "This CGL policy requires the insured to pay a $25,000 deductible per occurrence before coverage activates. The deductible reduces the $1,000,000 per-occurrence limit but does not affect the $2,000,000 aggregate limit, meaning multiple claims can accumulate toward the aggregate without additional deductible reductions."}
+{"chunk_id": "7cde98aa-d6f2-4d42-a8b5-b033e8e7cd5b", "chunk_index": 2, "line_end": 23, "line_start": 17, "one_liner": "This CGL policy (GNM-2024-88714) excludes pollution liability, professional services errors and omissions, and cyber/data breach costs, which are covered under separate specialized policies. The policy provides $2M aggregate and $1M per-occurrence limits with a $25K deductible per occurrence."}

parsed/942c3ac5-467f-4b9f-92ca-cbbb2785c634/parsed.json

@@ -0,0 +1 @@
+{"mime": "text/markdown", "page_count": 1, "parser": "markdown", "raw": null, "sections": [{"bbox": [{"height": 5, "left": 0, "page": 1, "text": null, "top": 1, "type": "MarkdownHeading", "width": 0}], "metadata": {}, "page_end": 1, "page_start": 1, "section_type": null, "text": "# CGL Policy Summary — GNM-2024-88714\n\nThis Commercial General Liability policy covers bodily injury and\nproperty damage claims arising from premises operations. The aggregate\nlimit is $2,000,000 per policy period and $1,000,000 per occurrence.", "title": "CGL Policy Summary — GNM-2024-88714"}, {"bbox": [{"height": 5, "left": 0, "page": 1, "text": null, "top": 7, "type": "MarkdownHeading", "width": 0}], "metadata": {}, "page_end": 1, "page_start": 1, "section_type": null, "text": "## Deductible\n\nThe insured pays a $25,000 deductible per occurrence before coverage\napplies. The deductible erodes the per-occurrence limit but not the\naggregate.", "title": "Deductible"}, {"bbox": [{"height": 5, "left": 0, "page": 1, "text": null, "top": 13, "type": "MarkdownHeading", "width": 0}], "metadata": {}, "page_end": 1, "page_start": 1, "section_type": null, "text": "## Exclusions\n\n* Pollution liability (handled by separate environmental policy).\n* Professional services E&O.\n* Cyber events and data breach response costs.", "title": "Exclusions"}]}

sagalog.toml

@@ -1,7 +1,7 @@
 # sagalog path-space metadata (committed to git).
 
 [remote]
-url = ""
+url = "https://***@git.sagalog.huscarl.io/sagalog/harvey-extract.git"
 
 [embedder]
 id = "voyage-voyage-4"