okb 1.1.0a0__py3-none-any.whl → 1.1.1__py3-none-any.whl

okb/migrations/0008.enrichment.sql

@@ -0,0 +1,46 @@
+-- LLM enrichment for document annotation (TODOs and entities)
+-- depends: 0006.llm-cache
+
+-- Track enrichment state on documents
+ALTER TABLE documents ADD COLUMN IF NOT EXISTS enriched_at TIMESTAMPTZ;
+ALTER TABLE documents ADD COLUMN IF NOT EXISTS enrichment_version INTEGER;
+
+-- Index for "needs enrichment" queries
+CREATE INDEX IF NOT EXISTS idx_documents_needs_enrichment
+    ON documents(enriched_at) WHERE enriched_at IS NULL;
+
+-- Pending entity suggestions (before approval)
+CREATE TABLE IF NOT EXISTS pending_entities (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    source_document_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    entity_name TEXT NOT NULL,
+    entity_type TEXT NOT NULL,  -- person, project, technology, concept, organization
+    aliases JSONB DEFAULT '[]',
+    description TEXT,
+    mentions JSONB DEFAULT '[]',  -- Context snippets from source document
+    confidence REAL,
+    status TEXT DEFAULT 'pending',  -- pending, approved, rejected
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    reviewed_at TIMESTAMPTZ
+);
+
+CREATE INDEX IF NOT EXISTS idx_pending_entities_status ON pending_entities(status);
+CREATE INDEX IF NOT EXISTS idx_pending_entities_source ON pending_entities(source_document_id);
+CREATE INDEX IF NOT EXISTS idx_pending_entities_type ON pending_entities(entity_type);
+
+-- Entity references (links entity documents to source documents)
+-- When an entity is approved, it becomes a document with source_path like okb://entity/person/john-smith
+-- This table tracks which documents mention each entity
+CREATE TABLE IF NOT EXISTS entity_refs (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    entity_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    document_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    mention_text TEXT NOT NULL,
+    context TEXT,  -- Surrounding text for context
+    confidence REAL,
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    UNIQUE(entity_id, document_id, mention_text)
+);
+
+CREATE INDEX IF NOT EXISTS idx_entity_refs_entity ON entity_refs(entity_id);
+CREATE INDEX IF NOT EXISTS idx_entity_refs_document ON entity_refs(document_id);

okb/migrations/0009.entity-consolidation.sql

@@ -0,0 +1,120 @@
+-- Entity consolidation: deduplication, cross-doc detection, clustering, relationships
+-- depends: 0008.enrichment
+
+-- Canonical mappings: alias text -> entity document
+-- Used for deduplication and alias resolution
+CREATE TABLE IF NOT EXISTS entity_aliases (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    alias_text TEXT NOT NULL,
+    entity_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    confidence REAL,  -- How confident we are this alias belongs to entity
+    source TEXT DEFAULT 'manual',  -- 'manual', 'merge', 'extraction'
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    UNIQUE(alias_text, entity_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_entity_aliases_text ON entity_aliases(LOWER(alias_text));
+CREATE INDEX IF NOT EXISTS idx_entity_aliases_entity ON entity_aliases(entity_id);
+
+-- Proposed entity merges awaiting user confirmation
+CREATE TABLE IF NOT EXISTS pending_entity_merges (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    canonical_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    duplicate_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    confidence REAL NOT NULL,  -- How confident we are these are the same
+    reason TEXT,  -- Why we think they're the same ("embedding_similarity", "alias_match", "llm")
+    detected_at TIMESTAMPTZ DEFAULT NOW(),
+    status TEXT DEFAULT 'pending',  -- 'pending', 'approved', 'rejected'
+    reviewed_at TIMESTAMPTZ,
+    UNIQUE(canonical_id, duplicate_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_pending_merges_status ON pending_entity_merges(status);
+CREATE INDEX IF NOT EXISTS idx_pending_merges_confidence ON pending_entity_merges(confidence DESC);
+
+-- Entity-to-entity relationships
+CREATE TABLE IF NOT EXISTS entity_relationships (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    source_entity_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    target_entity_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    relationship_type TEXT NOT NULL,  -- 'works_for', 'uses', 'belongs_to', 'related_to'
+    confidence REAL,
+    source TEXT DEFAULT 'extraction',  -- 'extraction', 'manual'
+    context TEXT,  -- Supporting context for the relationship
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    UNIQUE(source_entity_id, target_entity_id, relationship_type)
+);
+
+CREATE INDEX IF NOT EXISTS idx_entity_rel_source ON entity_relationships(source_entity_id);
+CREATE INDEX IF NOT EXISTS idx_entity_rel_target ON entity_relationships(target_entity_id);
+CREATE INDEX IF NOT EXISTS idx_entity_rel_type ON entity_relationships(relationship_type);
+
+-- Topic clusters group related entities and documents
+CREATE TABLE IF NOT EXISTS topic_clusters (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    name TEXT NOT NULL,
+    description TEXT,
+    centroid vector(768),  -- Cluster centroid embedding
+    member_count INTEGER DEFAULT 0,
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    updated_at TIMESTAMPTZ DEFAULT NOW()
+);
+
+CREATE INDEX IF NOT EXISTS idx_topic_clusters_centroid ON topic_clusters
+    USING hnsw (centroid vector_cosine_ops);
+
+-- Cluster membership: entities and documents can belong to clusters
+CREATE TABLE IF NOT EXISTS topic_cluster_members (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    cluster_id UUID NOT NULL REFERENCES topic_clusters(id) ON DELETE CASCADE,
+    document_id UUID NOT NULL REFERENCES documents(id) ON DELETE CASCADE,
+    distance REAL,  -- Distance from cluster centroid
+    is_entity BOOLEAN DEFAULT FALSE,  -- True if document is an entity
+    added_at TIMESTAMPTZ DEFAULT NOW(),
+    UNIQUE(cluster_id, document_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_cluster_members_cluster ON topic_cluster_members(cluster_id);
+CREATE INDEX IF NOT EXISTS idx_cluster_members_document ON topic_cluster_members(document_id);
+
+-- Proposed cluster merges awaiting confirmation
+CREATE TABLE IF NOT EXISTS pending_cluster_merges (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    primary_cluster_id UUID NOT NULL REFERENCES topic_clusters(id) ON DELETE CASCADE,
+    secondary_cluster_id UUID NOT NULL REFERENCES topic_clusters(id) ON DELETE CASCADE,
+    similarity REAL NOT NULL,  -- How similar the clusters are
+    status TEXT DEFAULT 'pending',  -- 'pending', 'approved', 'rejected'
+    detected_at TIMESTAMPTZ DEFAULT NOW(),
+    reviewed_at TIMESTAMPTZ,
+    UNIQUE(primary_cluster_id, secondary_cluster_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_pending_cluster_merges_status ON pending_cluster_merges(status);
+
+-- Cross-document entity candidates: detected mentions not yet extracted as entities
+CREATE TABLE IF NOT EXISTS cross_doc_entity_candidates (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    text TEXT NOT NULL,  -- The mention text (normalized)
+    document_ids UUID[] NOT NULL,  -- Array of document IDs containing this mention
+    document_count INTEGER NOT NULL,  -- Number of documents (for quick filtering)
+    sample_contexts JSONB DEFAULT '[]',  -- Sample text contexts where it appears
+    suggested_type TEXT,  -- Suggested entity type
+    confidence REAL,
+    status TEXT DEFAULT 'pending',  -- 'pending', 'approved', 'rejected', 'exists'
+    created_at TIMESTAMPTZ DEFAULT NOW(),
+    reviewed_at TIMESTAMPTZ,
+    UNIQUE(text)
+);
+
+CREATE INDEX IF NOT EXISTS idx_cross_doc_status ON cross_doc_entity_candidates(status);
+CREATE INDEX IF NOT EXISTS idx_cross_doc_count ON cross_doc_entity_candidates(document_count DESC);
+
+-- Track consolidation runs
+CREATE TABLE IF NOT EXISTS consolidation_runs (
+    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+    run_type TEXT NOT NULL,  -- 'dedup', 'cross_doc', 'cluster', 'relationship', 'full'
+    started_at TIMESTAMPTZ DEFAULT NOW(),
+    completed_at TIMESTAMPTZ,
+    stats JSONB DEFAULT '{}',  -- Run statistics
+    error TEXT  -- Error message if failed
+);

okb/migrations/0010.token-id.sql

@@ -0,0 +1,7 @@
+-- Add ID column to tokens for easier revocation
+-- depends: 0009.entity-consolidation
+
+ALTER TABLE tokens ADD COLUMN IF NOT EXISTS id SERIAL;
+
+-- Create index for ID lookups
+CREATE INDEX IF NOT EXISTS tokens_id_idx ON tokens(id);

okb/modal_llm.py

@@ -2,20 +2,34 @@
 Modal-based GPU LLM service for document classification.
 
 Provides on-demand GPU access for LLM inference using open models.
-Uses Llama 3.2 3B by default - fast and efficient for classification tasks.
 
 Usage:
+    # Deploy with default model (Phi-3)
     modal deploy modal_llm.py
 
+    # Deploy with specific model
+    OKB_LLM_MODEL=meta-llama/Llama-3.2-3B-Instruct modal deploy modal_llm.py
+
 Then call from Python:
     llm = modal.Cls.from_name("knowledge-llm", "LLM")()
     response = llm.complete.remote("Classify this document", system="You are a classifier")
 """
 
+import os
+
 import modal
 
 app = modal.App("knowledge-llm")
 
+# Model is set via environment variable at deploy time
+# Default to Phi-3 which doesn't require HuggingFace approval
+DEFAULT_MODEL = "microsoft/Phi-3-mini-4k-instruct"
+MODEL_ID = os.environ.get("OKB_LLM_MODEL", DEFAULT_MODEL)
+
+# GPU type - L4 recommended for speed/cost balance
+DEFAULT_GPU = "L4"
+GPU_TYPE = os.environ.get("OKB_MODAL_GPU", DEFAULT_GPU)
+
 # Container image with transformers and torch
 llm_image = (
     modal.Image.debian_slim(python_version="3.11")
@@ -24,17 +38,19 @@ llm_image = (
         "torch>=2.0.0",
         "accelerate>=0.27.0",
         "bitsandbytes>=0.42.0",  # For quantization
+        "hf_transfer",  # Fast downloads
     )
-    .env({"HF_HUB_ENABLE_HF_TRANSFER": "1"})
+    .env({
+        "HF_HUB_ENABLE_HF_TRANSFER": "1",
+        "OKB_LLM_MODEL": MODEL_ID,
+        "OKB_MODAL_GPU": GPU_TYPE,
+    })
 )
 
-# Default model - Llama 3.2 3B is fast and good for classification
-DEFAULT_MODEL = "meta-llama/Llama-3.2-3B-Instruct"
-
 
 @app.cls(
     image=llm_image,
-    gpu="T4",  # T4 is sufficient for 3B model with quantization
+    gpu=GPU_TYPE,
     timeout=300,
     scaledown_window=300,  # Keep warm for 5 min
     retries=1,
@@ -42,14 +58,16 @@ DEFAULT_MODEL = "meta-llama/Llama-3.2-3B-Instruct"
 class LLM:
     """GPU-accelerated LLM for document classification."""
 
-    model_id: str = DEFAULT_MODEL
-
     @modal.enter()
     def load_model(self):
         """Load model once when container starts."""
+        import os
+
         import torch
         from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 
+        # Read model from environment (set at deploy time)
+        self.model_id = os.environ.get("OKB_LLM_MODEL", "microsoft/Phi-3-mini-4k-instruct")
         print(f"Loading model: {self.model_id}")
 
         # Use 4-bit quantization for memory efficiency

okb/plugins/sources/github.py

@@ -95,11 +95,11 @@ class GitHubSource:
               token: ${GITHUB_TOKEN}
 
     Usage:
-        lkb sync run github --repo owner/repo              # README + docs/ (default)
-        lkb sync run github --repo owner/repo --source     # All source files
-        lkb sync run github --repo owner/repo --issues     # Include issues
-        lkb sync run github --repo owner/repo --prs        # Include PRs
-        lkb sync run github --repo owner/repo --wiki       # Include wiki
+        okb sync run github --repo owner/repo              # README + docs/ (default)
+        okb sync run github --repo owner/repo --source     # All source files
+        okb sync run github --repo owner/repo --issues     # Include issues
+        okb sync run github --repo owner/repo --prs        # Include PRs
+        okb sync run github --repo owner/repo --wiki       # Include wiki
     """
 
     name = "github"

okb/tokens.py

@@ -22,6 +22,7 @@ from psycopg.rows import dict_row
 class TokenInfo:
     """Information about a token."""
 
+    id: int
     token_hash: str
     database: str
     permissions: str  # 'ro' or 'rw'
@@ -143,7 +144,7 @@ def list_tokens(db_url: str) -> list[TokenInfo]:
         with psycopg.connect(db_url, row_factory=dict_row) as conn:
             results = conn.execute(
                 """
-                SELECT token_hash, permissions, description, created_at, last_used_at
+                SELECT id, token_hash, permissions, description, created_at, last_used_at
                 FROM tokens
                 ORDER BY created_at DESC
                 """
@@ -151,6 +152,7 @@ def list_tokens(db_url: str) -> list[TokenInfo]:
 
             return [
                 TokenInfo(
+                    id=r["id"],
                     token_hash=r["token_hash"],
                     database=db_name,
                     permissions=r["permissions"],
@@ -171,7 +173,7 @@ def delete_token(db_url: str, token_or_prefix: str) -> bool:
 
     Args:
         db_url: Database connection URL
-        token_or_prefix: Full token or token prefix (e.g., 'lkb_personal_ro')
+        token_or_prefix: Full token or token prefix (e.g., 'okb_personal_ro')
 
     Returns:
         True if token was deleted, False if not found
@@ -199,6 +201,25 @@ def delete_token(db_url: str, token_or_prefix: str) -> bool:
         return False
 
 
+def delete_token_by_id(db_url: str, token_id: int) -> bool:
+    """Delete a token by its ID.
+
+    Args:
+        db_url: Database connection URL
+        token_id: Token ID from the tokens table
+
+    Returns:
+        True if token was deleted, False if not found
+    """
+    with psycopg.connect(db_url) as conn:
+        result = conn.execute(
+            "DELETE FROM tokens WHERE id = %s RETURNING id",
+            (token_id,),
+        ).fetchone()
+        conn.commit()
+        return result is not None
+
+
 def verify_token(token: str, get_db_url_fn) -> TokenInfo | None:
     """Verify a token and return its info if valid.
 
@@ -225,7 +246,7 @@ def verify_token(token: str, get_db_url_fn) -> TokenInfo | None:
         with psycopg.connect(db_url, row_factory=dict_row) as conn:
             result = conn.execute(
                 """
-                SELECT token_hash, permissions, description, created_at, last_used_at
+                SELECT id, token_hash, permissions, description, created_at, last_used_at
                 FROM tokens
                 WHERE token_hash = %s
                 """,
@@ -243,6 +264,7 @@ def verify_token(token: str, get_db_url_fn) -> TokenInfo | None:
             conn.commit()
 
             return TokenInfo(
+                id=result["id"],
                 token_hash=result["token_hash"],
                 database=database,
                 permissions=result["permissions"],

{okb-1.1.0a0.dist-info → okb-1.1.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: okb
-Version: 1.1.0a0
+Version: 1.1.1
 Summary: Personal knowledge base with semantic search for LLMs
 Requires-Python: >=3.11
 Classifier: Programming Language :: Python :: 3
@@ -85,6 +85,8 @@ okb ingest ~/notes ~/docs
 | `okb db start` | Start pgvector database container |
 | `okb db stop` | Stop database container |
 | `okb db status` | Show database status |
+| `okb db migrate [name]` | Apply pending migrations (optionally for specific db) |
+| `okb db list` | List configured databases |
 | `okb db destroy` | Remove container and volume (destructive) |
 | `okb ingest <paths>` | Ingest documents into knowledge base |
 | `okb ingest <paths> --local` | Ingest using local GPU/CPU embedding (no Modal) |
@@ -93,10 +95,11 @@ okb ingest ~/notes ~/docs
 | `okb watch <paths>` | Watch directories for changes |
 | `okb config init` | Create default config file |
 | `okb config show` | Show current configuration |
+| `okb config path` | Print config file path |
 | `okb modal deploy` | Deploy GPU embedder to Modal |
 | `okb token create` | Create API token for HTTP server |
 | `okb token list` | List tokens for a database |
-| `okb token revoke` | Revoke an API token |
+| `okb token revoke [TOKEN] --id <n>` | Revoke token by full value or ID |
 | `okb sync list` | List available API sources (plugins) |
 | `okb sync list-projects <source>` | List projects from source (for config) |
 | `okb sync run <sources>` | Sync data from external APIs |
@@ -108,6 +111,18 @@ okb ingest ~/notes ~/docs
 | `okb llm status` | Show LLM config and connectivity |
 | `okb llm deploy` | Deploy Modal LLM for open model inference |
 | `okb llm clear-cache` | Clear LLM response cache |
+| `okb enrich run` | Extract TODOs and entities from documents |
+| `okb enrich run --dry-run` | Show what would be enriched |
+| `okb enrich pending` | List entities awaiting review |
+| `okb enrich approve <id>` | Approve a pending entity |
+| `okb enrich reject <id>` | Reject a pending entity |
+| `okb enrich analyze` | Analyze database and update description/topics |
+| `okb enrich consolidate` | Run entity consolidation (duplicates, clusters) |
+| `okb enrich merge-proposals` | List pending merge proposals |
+| `okb enrich approve-merge <id>` | Approve an entity merge |
+| `okb enrich reject-merge <id>` | Reject an entity merge |
+| `okb enrich clusters` | List topic clusters |
+| `okb enrich relationships` | List entity relationships |
 
 
 ## Configuration
@@ -142,7 +157,7 @@ chunking:
 Use `--db <name>` to target a specific database with any command.
 
 Environment variables override config file settings:
-- `KB_DATABASE_URL` - Database connection string
+- `OKB_DATABASE_URL` - Database connection string
 - `OKB_DOCKER_PORT` - Docker port mapping
 - `OKB_CONTAINER_NAME` - Docker container name
 
@@ -163,7 +178,7 @@ Merge: scalars replace, lists extend, dicts deep-merge.
 
 ### LLM Integration (Optional)
 
-Enable LLM-based document classification and filtering:
+Enable LLM-based document classification, filtering, and enrichment:
 
 ```yaml
 llm:
@@ -179,11 +194,25 @@ llm:
 | `claude` | `export ANTHROPIC_API_KEY=...` | ~$0.25/1M tokens |
 | `modal` | `okb llm deploy` | ~$0.02/min GPU |
 
-For Modal (no API key needed):
+**Modal LLM Setup** (no API key needed, runs on Modal's GPUs):
+
 ```yaml
 llm:
   provider: modal
-  model: meta-llama/Llama-3.2-3B-Instruct
+  model: microsoft/Phi-3-mini-4k-instruct  # Recommended: no gating
+```
+
+Non-gated models (work immediately):
+- `microsoft/Phi-3-mini-4k-instruct` - Good quality, 4K context
+- `Qwen/Qwen2-1.5B-Instruct` - Smaller/faster
+
+Gated models (require HuggingFace approval + token):
+- `meta-llama/Llama-3.2-3B-Instruct` - Requires accepting license at HuggingFace
+- Setup: `modal secret create huggingface HF_TOKEN=hf_...`
+
+Deploy after configuring:
+```bash
+okb llm deploy
 ```
 
 **Pre-ingest filtering** - skip low-value content during sync:
@@ -197,6 +226,36 @@ plugins:
         action_on_skip: discard  # or "archive"
 ```
 
+### Document Enrichment
+
+Extract TODOs and entities (people, projects, technologies) from documents using LLM:
+
+```bash
+okb enrich run                      # Enrich un-enriched documents
+okb enrich run --dry-run            # Preview what would be enriched
+okb enrich run --source-type markdown  # Only markdown files
+okb enrich run --query "meeting"    # Filter by semantic search
+```
+
+Entities are created as pending suggestions for review:
+```bash
+okb enrich pending                  # List pending entities
+okb enrich approve <id>             # Approve → creates entity document
+okb enrich reject <id>              # Reject → hidden from future suggestions
+```
+
+Configure enrichment behavior:
+```yaml
+enrichment:
+  enabled: true
+  extract_todos: true
+  extract_entities: true
+  auto_create_todos: true       # TODOs created immediately
+  auto_create_entities: false   # Entities go to pending review
+  min_confidence_todo: 0.7
+  min_confidence_entity: 0.8
+```
+
 CLI commands:
 ```bash
 okb llm status              # Show config and connectivity
@@ -234,14 +293,20 @@ okb token create --db default -d "Claude Code"
 okb serve --http --host 0.0.0.0 --port 8080
 ```
 
-Then configure Claude Code to connect via SSE:
+The server uses Streamable HTTP transport (RFC 9728 compliant):
+- `POST /mcp` - Send JSON-RPC messages, receive SSE response
+- `GET /mcp` - Establish SSE connection for server notifications
+- `DELETE /mcp` - Terminate session
+- `/sse` is an alias for `/mcp` for backward compatibility
+
+Configure your MCP client to connect:
 
 ```json
 {
   "mcpServers": {
     "knowledge-base": {
       "type": "sse",
-      "url": "http://localhost:8080/sse",
+      "url": "http://localhost:8080/mcp",
       "headers": {
         "Authorization": "Bearer okb_default_rw_a1b2c3d4e5f6g7h8"
       }
@@ -269,6 +334,20 @@ Then configure Claude Code to connect via SSE:
 | `add_todo` | Create a TODO item in the knowledge base |
 | `trigger_sync` | Sync API sources (Todoist, GitHub, Dropbox Paper) |
 | `trigger_rescan` | Check indexed files for changes and re-ingest |
+| `list_sync_sources` | List available API sync sources with status |
+| `enrich_document` | Run LLM enrichment to extract TODOs/entities |
+| `list_pending_entities` | List entities awaiting review |
+| `approve_entity` | Approve a pending entity |
+| `reject_entity` | Reject a pending entity |
+| `analyze_knowledge_base` | Analyze content and generate description/topics |
+| `find_entity_duplicates` | Find potential duplicate entities |
+| `merge_entities` | Merge duplicate entities |
+| `list_pending_merges` | List pending merge proposals |
+| `approve_merge` | Approve a merge proposal |
+| `reject_merge` | Reject a merge proposal |
+| `get_topic_clusters` | Get topic clusters from consolidation |
+| `get_entity_relationships` | Get relationships between entities |
+| `run_consolidation` | Run full entity consolidation pipeline |
 
 ## Contextual Chunking
 
@@ -291,6 +370,10 @@ project: student-app
 category: backend
 ---
 
+# Your Document Title
+
+Content here...
+```
 
 ## Plugin System
 

{okb-1.1.0a0.dist-info → okb-1.1.1.dist-info}/RECORD

@@ -1,16 +1,25 @@
 okb/__init__.py,sha256=2yaWIYQbho7N2O2zwTn3ZH11b8b3SaoDVlxluVTqwy4,92
-okb/cli.py,sha256=y8Vr9Scy7PyAtgrCb2yIsN3kRvhwUvxpnpiF6RVV_MA,47735
-okb/config.py,sha256=DKmX2fgteGdh0QMsA-Immu-mZcvLjHWeB8HIf9rcM5o,22898
+okb/cli.py,sha256=8v_SaXFOrJYrCPSr6JgIqqzFHHYpPlHvan4CLaYUDTs,88639
+okb/config.py,sha256=vKDC6b6Tm3_XZzvn7nA9WlGCWzCT8vtV9AvLes02YW8,28562
 okb/data/init.sql,sha256=QpsicUN7PQ7d8zyOCRNChOu5XKdUVC3xySlRDPyKSN8,2728
-okb/http_server.py,sha256=jcpNWB1aGtcHE7h0U4gCxA4lZyqWHGgsiArv7DyPSZw,20595
+okb/http_server.py,sha256=QdF7KDWrgvbEWXkm8nrBansKN7YbGE28yjEXC9bEXJc,30139
 okb/ingest.py,sha256=D5plxCC2tQXZenMNUa482dUDqsyuaq2APAQqaIgRAqU,54505
 okb/llm/__init__.py,sha256=4jelqgXvF-eEPyLCuAmcxagN0H923wI9pBJJZKv4r0E,2368
+okb/llm/analyze.py,sha256=BKW308AtjWStZcZiMKaRqFmQsuTclp3Qp3W4nsdw4vk,18569
 okb/llm/base.py,sha256=gOm7zBiNdHrj7xxJfpb-4qZdYxWM0lA0vKfrBStO60E,2279
 okb/llm/cache.py,sha256=rxRPMNBtP336MSpGWA8F7rDZnF0O2RM3rEsNtoxS0Zk,6142
+okb/llm/consolidate.py,sha256=TjXBWzzlJrX3_z0CysrFAxclp1K7XGgMb_SI5hdK7Y8,23178
+okb/llm/enrich.py,sha256=Yc09xvXynuiCATK451J5sBQ9lkPPllio8vAPYOEVDFo,23133
+okb/llm/extractors/__init__.py,sha256=8sHgtgtydF4CBFiIOfURYcyunVxK770qp_aMz4jxnFU,317
+okb/llm/extractors/base.py,sha256=p_PIMdUoEw8-C5jG8TcbG73jegE5ovmIGbbbdDrRFVs,1250
+okb/llm/extractors/cross_doc.py,sha256=nw1-nHSHKxGlL95rfHFkStdbzpy4H3eRclyUyn2YE1A,18114
+okb/llm/extractors/dedup.py,sha256=TVDBhNX7rPmpipO2WckAhe_Oj1LmSRPce-EceZ6FdbM,16562
+okb/llm/extractors/entity.py,sha256=woIF5AeXD2dmFDoiOVCPgMUVWQPYHOmg1KE-ZRB3o0E,9925
+okb/llm/extractors/todo.py,sha256=LdAyioQ5fZ-KJC_W_-CAU8OlMvju3Oci7E7aueoTnpI,4345
 okb/llm/filter.py,sha256=y20bc3vHtp5gj7T7AhsJ45ZkAkBgztj6WPjsVAmvEeo,5447
-okb/llm/providers.py,sha256=AdVw9FFgv58-KJEfXv9JqWlkxBl-LcRWOao95CsjqWA,9718
+okb/llm/providers.py,sha256=SpbEpJYSQHc43qyvBdv2IuUNKsf0_NfDmnooNM1tzHc,9838
 okb/local_embedder.py,sha256=zzjBUFp4IH2xsvKyKjKZyX9dJuE_3PDMHMwpyRYSISQ,2098
-okb/mcp_server.py,sha256=BnMxyGf524sK-8CYPyL3ZM_DEqWFsXpF7_66xj3-Ecs,59407
+okb/mcp_server.py,sha256=IUEGo4xmo2HhvKHhgeADqvNxICk4dBTX-6BNSmi8fkY,98610
 okb/migrate.py,sha256=2faYL-SHiQCkGXpTUlBFMCj0B-6JYCHqZl9u6vOlui8,1693
 okb/migrations/0001.initial-schema.sql,sha256=0s5pj9Ad6f0u_mxODAM_-DbDI3aI37Wdu5XMPAzAIqw,2577
 okb/migrations/0002.sync-state.sql,sha256=w34dOA9sPg60NMS1aHvOhORff1k_Di9cO2ghwVQSPHU,696
@@ -18,20 +27,23 @@ okb/migrations/0003.structured-fields.sql,sha256=rPCSrdtotCoRpOfjHf1Ifx0pfizpYS9
 okb/migrations/0004.tokens.sql,sha256=VtcLfA1_SVVQLkEKZ-av_93Fg0ksVWLm1tlR7nJXoaQ,448
 okb/migrations/0005.database-metadata.sql,sha256=0X4LyuUUX34s3ph2C70FnBBau5HEBwR4xyY-hwPEX90,709
 okb/migrations/0006.llm-cache.sql,sha256=azjPpj00WH_8tx4JI8PJKZ1AOAJEhbkneVvYa3ZRZ1w,493
+okb/migrations/0008.enrichment.sql,sha256=pkKtyuFgRiPhZ7YAyYr6yihT2R-Tx64hE18iqphqFAw,2160
+okb/migrations/0009.entity-consolidation.sql,sha256=Q2TBHJqMa1fvAqaopa20KdJCIxqldDAeOPGC5YiuzxY,5800
+okb/migrations/0010.token-id.sql,sha256=odFPvY3z63f6Zfx8j5qBj2LobvPafpT2hDi_MwBBf6Q,231
 okb/modal_embedder.py,sha256=V1cpGWrtEo1MGkrD9Nc-5wRmf9e7IwKPsQj7nVuLlyg,3469
-okb/modal_llm.py,sha256=4rYE3VZ_T09HXCgTIYFLu1s_C2FRC9y4dgMUGqJuO2M,5368
+okb/modal_llm.py,sha256=NpCykbjlhUUdQNQ4NMLu-uEBFJbYwJDV2NdnwUZeG84,5903
 okb/plugins/__init__.py,sha256=50LNAH4bvfIw5CHT82sknGjdCldQ-4ds0wxo1zM9E2k,324
 okb/plugins/base.py,sha256=6TIN1UIItmuIsP4NDJhuRMH0ngKkQiGmtHTeYj1K8OU,3171
 okb/plugins/registry.py,sha256=fN7NfoOaRnMyXSWT2srd6vEr4riJjmncQFfberf0IE8,3741
 okb/plugins/sources/__init__.py,sha256=n58rAbcJC45JbofUY6IA526rSRjkYn4_tGjWma3TOUI,214
 okb/plugins/sources/dropbox_paper.py,sha256=Oi59NbJGQrwjE2Xhcinc2InKRc27Gdg7l8xVTbKLkI8,7493
-okb/plugins/sources/github.py,sha256=ozdTZPkU8h2-ZIx5o1FB58QBZ6P0eoVntluWL3vG87I,16309
+okb/plugins/sources/github.py,sha256=YlwsY2hRKlGTyHL5L4vCC1TP-b6WJzEYbPDEXnB4y-I,16309
 okb/plugins/sources/todoist.py,sha256=B22tKYFZhuDhZHhpRdGWDGho9y7FBNgGlI1g2nf13-8,8849
 okb/rescan.py,sha256=dVdQEkVUjsrtOKAGZc0LC2uwcnkjB8hn2SOVWHnY-R8,8396
 okb/scripts/__init__.py,sha256=HPp8YCtIeo9XMOtOGCtntiwYr9eCxAJ1MF9Lo9WVzUA,53
 okb/scripts/watch.py,sha256=b8oGPTN3flNdNQJETeqQ1RNZ8U1LiKvHntLwvHRIviA,6354
-okb/tokens.py,sha256=JJ1C-mvtnT2O0cmjSu57PI9Nt53Sl9DqbmPuLnHlN6g,8043
-okb-1.1.0a0.dist-info/METADATA,sha256=IhNkQv-lucqtYIaXcNfxkKFkSD5Avo3Vy5buDbXHELo,10578
-okb-1.1.0a0.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
-okb-1.1.0a0.dist-info/entry_points.txt,sha256=YX6b8BlV9sSAXrneoIm3dkXtRcgHhSzbDaOpJ0yCKRs,230
-okb-1.1.0a0.dist-info/RECORD,,
+okb/tokens.py,sha256=3Of_PwNCTTexXC3d-EAiPjLdsbyk2F_dTeY30O3mqp8,8635
+okb-1.1.1.dist-info/METADATA,sha256=n_hhat8q-7KJGyfcDhCqA_tqLLWKKgRIDgLpOpZ2IEo,14113
+okb-1.1.1.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+okb-1.1.1.dist-info/entry_points.txt,sha256=YX6b8BlV9sSAXrneoIm3dkXtRcgHhSzbDaOpJ0yCKRs,230
+okb-1.1.1.dist-info/RECORD,,