| df0a485cf6dffdef… | A DraftCXU has a `knowledge_type` enum field with values `axiom`, `derived`, or `prescribed` so that the unit’s knowledge category is explicitly classified. | active | contextual | yes | |
| f4818ac85a45dc56… | A DraftCXU has a `supporting_quotes` field that is a list containing 3 to 6 verbatim source quotes so that each claim is backed by direct textual evidence. | active | contextual | yes | |
| c6e8ffad81fdbd9b… | A DraftCXU has a `claim` field of type string that stores a standalone assertion between 15 and 70 words so that each CXU expresses one self-contained statement. | active | contextual | yes | |
| 71f290f2483ace22… | A DraftCXU has a `cxu_id` field of type string formatted as `1220` plus a SHA-256 hex value with 68 characters so that each draft context unit has a deterministic identifier. | active | contextual | yes | |
| 9361c149e3726abc… | Named relationships are managed through a set-relationship registry with a lifecycle of pending, active, and inactive, plus usage tracking and co-occurrence discovery so that relationship definitions can be governed and analyzed over time. | active | contextual | yes | |
| 4b12327db0f93675… | The extraction system supports the profiles `general` and `pharma-manufacturing` so that extraction behavior can be tailored to different domains. | active | contextual | yes | |
| 1e3bf77f3369685e… | The extraction system uses lenses for facts, definitions, procedures, requirements, and specifications so that different knowledge forms can be extracted in a targeted way. | active | contextual | yes | |
| f06041cdf6dabf9a… | Extraction follows a two-phase process of chunking followed by parallel lens extraction so that source material is first segmented and then analyzed concurrently through multiple lenses. | active | contextual | yes | |
| 72741c5a15390d61… | The coordinated CORTEX agent layer includes an AmbientAgent for proactive gap detection so that missing knowledge can be identified without waiting for explicit requests. | active | contextual | yes | |
| a0089ad78ad903f0… | The coordinated CORTEX agent layer includes a LearningAgent that uses TEMPR memory so that the system can maintain and use learned memory structures. | active | contextual | yes | |
| 2df4bdde02ae79a5… | The CORTEX agent layer is coordinated and includes a RetrievalAgent for Q-value reranking, cross-encoder scoring, and two-level retrieval so that retrieval quality is improved through layered ranking logic. | active | contextual | yes | |
| 7be8ebe82d9248a3… | KAG retrieval uses a 3-strategy pipeline consisting of template Cypher, edge-filter Cypher, and Text2Cypher free search, and the results are fused with Q-value weighted RRF so that multiple retrieval methods contribute to final ranking. | active | contextual | yes | |
| 4b2fe2902569ce30… | Sets are named many-to-many CXU groupings, and the UNIVERSAL set is always in scope so that CXUs can be organized into reusable collections with one default global set. | active | contextual | yes | |
| bbe2d14a6fd77d1c… | The CXU lifecycle progresses through DRAFT, ACTIVE, INACTIVE, and VERSIONED states, with an approval workflow required for REGULATED and ABSOLUTE classifications so that sensitive classifications receive governance before activation. | active | contextual | yes | |
| 597980e90a8c27ac… | A CXU is an atomic knowledge unit containing a claim, supporting quotes, and metadata, and it uses a content-addressable ID based on `1220` plus SHA-256 so that each unit is self-contained and uniquely addressable. | active | contextual | yes | |
| b0ae50585ca64a8c… | Lifecycle endpoints support suggestions, approvals, and lifecycle state transitions so that CXU lifecycle management actions can be performed through the API. | active | contextual | yes | |
| ce6864368249fc4c… | The FastAPI API surface exposes routes including `/cxu`, `/healthz`, `/readyz`, `/api/v1/context`, `/api/v1/rank0/*`, `/api/v1/sets/*`, `/api/v1/cortex/*`, `/api/v1/reports/*`, `/extract`, `/analyze`, and `/documents` so that platform capabilities are available over HTTP endpoints. | active | contextual | yes | |
| f1030f99a262fce4… | The `context_engine.kag` package contains the Neo4j graph backend, Qdrant vector backend, and embedding generator so that graph, vector, and embedding operations are grouped in one package. | active | contextual | yes | |
| ea46da1982c49471… | The `context_engine.services` package contains the query processor with a 3-strategy pipeline and the Text2Cypher service so that query execution and graph-query generation are exposed as services. | active | contextual | yes | |
| 026486ab39ea2e7f… | The `context_engine.cortex` package contains CORTEX agents, the message bus, TEMPR memory, a reranker, and a set-relationship registry so that coordinated agent behavior and memory functions are implemented there. | active | contextual | yes | |
| 903d2e3f3038eccc… | The `context_engine.extract` package performs LLM extraction through a LiteLLM proxy and Instructor for structured output so that extracted results can be generated in a typed format. | active | contextual | yes | |
| bfe7f969a110efe2… | The `context_engine.core` package contains CXU models, consolidation, and similarity detection so that core knowledge-unit structures and comparison logic are centralized there. | active | contextual | yes | |
| 2557d00ce49da6a0… | The extraction pipeline accesses LLMs through a LiteLLM proxy that is OpenAI-compatible so that model routing can be handled behind a unified interface. | active | contextual | yes | |
| 6ae96d59bfe3ccc3… | NATS provides L2 agent coordination in the Context Engine using nats-py so that agents can coordinate through messaging infrastructure. | active | contextual | yes | |
| 8e8e8ef9cdb8aa99… | Redis functions as the cache backend in the Context Engine using the redis client so that frequently accessed data can be served more efficiently. | active | contextual | yes | |
| c52664a76a88b6d4… | Neo4j provides the knowledge graph and TEMPR memory backend in the Context Engine using neo4j-driver so that graph relationships and memory structures are stored there. | active | contextual | yes | |
| aa12298129d79ddb… | PostgreSQL serves as the source of truth and provides JSONB and full-text search capabilities in the Context Engine using asyncpg so that primary structured storage is centralized there. | active | contextual | yes | |
| b2f9f85ff70e9a75… | The Context Engine is a multi-backend Knowledge-Augmented Graph system with five storage and messaging backends so that knowledge, vectors, cache, and coordination are distributed across specialized services. | active | contextual | yes | |
| fe3e53e7fb8c737c… | Tenant overlays do not contain platform code; they only configure platform behavior per tenant so that customization remains separate from core implementation. | active | contextual | yes | |
| e77458d543f44160… | Each tenant overlay contains per-tenant agent definitions, workflow configurations, data connectors, and SQL templates so that platform behavior can be configured separately for each tenant. | active | contextual | yes | |
| 065523568480bf0a… | The `IN_SET` edge represents CxU to Set membership so that graph relationships can capture which sets include each context unit. | active | contextual | yes | |
| 09f7cbb0980830eb… | The `DERIVED_FROM` edge connects one CxU to another CxU as lineage so that derivation relationships between context units are preserved. | active | contextual | yes | |
| 2f8c653317cbc2ee… | The `PRODUCED` edge connects an AgentRun to a CxU so that the graph records which agent run generated a given context unit. | active | contextual | yes | |
| 4216ea19879ac0bb… | The `BELONGS_TO` edge connects a child node to its parent in the hierarchy so that hierarchical structure is represented explicitly in the graph. | active | contextual | yes | |
| 4283c19e70f831d6… | The `CONTEXT_FOR` edge connects a CxU to an entity and carries the properties `dimension`, `q_value`, and `is_rank0` so that contextual relevance and rank information are stored on the relationship. | active | contextual | yes | |
| 7b1ed75911dbf54a… | The Neo4j graph schema defines memory nodes as MemoryNode with relationships TEMPORAL_NEXT, SEMANTIC_SIMILAR, and REFERENCED so that memory structures can be linked by temporal, semantic, and reference relations. | active | contextual | yes | |
| 249e858468062e87… | The Neo4j graph schema defines agent nodes including AgentRun, ActionItem, and Analysis so that agent activities and outputs can be modeled as graph entities. | active | contextual | yes | |
| 20bbd77e0e04a746… | The Neo4j graph schema defines hierarchy nodes of Company, Pillar, Group, BusinessUnit, VirtualBusinessUnit, and CrossUnit so that organizational hierarchy can be represented in the graph. | active | contextual | yes | |
| 837bce7f5429f09f… | In the DraftCXU data model, `keywords` is a list so that searchable terms can be attached to each CXU. | active | contextual | yes | |
| e7b77792431af7fe… | In the DraftCXU data model, `parent_assumption_ids` is a list so that a CXU can reference related parent assumptions. | active | contextual | yes | |
| d36c23bf7e7f66a2… | In the DraftCXU data model, `is_assumption` is a boolean field so that a CXU can be marked as an assumption when applicable. | active | contextual | yes | |
| 0bf07d7f3c20db21… | In the DraftCXU data model, `claim_type` is an enum that includes values such as `definition`, `requirement`, and `procedure` so that the form of the claim can be classified. | active | contextual | yes | |
| 904cfb88d5044f55… | In the DraftCXU data model, `knowledge_type` is an enum with values `axiom`, `derived`, and `prescribed` so that each CXU is categorized by knowledge class. | active | contextual | yes | |
| 4ed18d32c69258f3… | In the DraftCXU data model, `supporting_quotes` is a list containing 3 to 6 verbatim source quotes so that each claim is backed by direct evidence from the source. | active | contextual | yes | |
| 44a8bed2c6e4c5d0… | In the DraftCXU data model, `claim` is a standalone assertion between 15 and 70 words so that each CXU expresses a self-contained statement of knowledge. | active | contextual | yes | |
| fe1fe2bb41762ae6… | In the DraftCXU data model, `cxu_id` is a string formatted as `1220` plus a SHA-256 hexadecimal value, producing 68 characters so that each CXU has a deterministic identifier. | active | contextual | yes | |
| a18ab47fb8a5c1ec… | Named Relationships are managed through a set-relationship registry with a lifecycle of pending, active, and inactive, along with usage tracking and co-occurrence discovery so that relationship definitions can be governed and analyzed. | active | contextual | yes | |
| 9ff00b935e3c50db… | Extraction is a two-phase process of chunking followed by parallel lens extraction, and it uses lenses for facts, definitions, procedures, requirements, and specifications plus profiles such as general and pharma-manufacturing so that extraction can be specialized by lens and domain. | active | contextual | yes | |
| 40f974b70e1ebf31… | CORTEX Agents form a coordinated agent layer including RetrievalAgent, LearningAgent, and AmbientAgent, with functions such as Q-value reranking, cross-encoder retrieval, TEMPR memory, and proactive gap detection so that agent responsibilities are specialized. | active | contextual | yes | |
| c52ed588b7a880a6… | KAG retrieval uses a three-strategy pipeline consisting of template Cypher, edge-filter Cypher, and Text2Cypher free search, and the results are fused using Q-value weighted RRF so that multiple retrieval methods are combined. | active | contextual | yes | |