| a1fd462f0fab35f8… | The Google Drive sync process should run the extraction pipeline on changed documents after detecting new or modified files. | active | contextual | yes | |
| ef5f4c23a83742b5… | A Google Drive sync job should be set up periodically to monitor both Drive folders for new or modified files as the first step of synchronization. | active | contextual | yes | |
| 3cbeaf2cbe1dbb06… | Once the knowledge graph is populated, the daily research task should query the context engine directly for platform knowledge instead of reading a flat file so that platform context becomes dynamic and always current. | active | contextual | yes | |
| 933301b551a9f8c8… | The context_search_entities MCP tool finds all entities related to a search term, with Patrick and code interpreter given as example queries. | active | contextual | yes | |
| ada1e740a7a05efc… | The context_feedback MCP tool records whether a retrieved CXU was helpful or unhelpful so that the system can feed Q-values. | active | contextual | yes | |
| fef45e953ba32e7c… | The context_publish_cxu MCP tool publishes a pre-formed CXU, with contract review given as an example source of such a CXU. | active | contextual | yes | |
| 19ea9bdd13d947eb… | The context_publish MCP tool publishes new CXUs from meeting notes or newly added documents into the internal knowledge graph. | active | contextual | yes | |
| 184a4343137b1326… | The context_graph_traverse MCP tool walks graph paths such as Client to Projects to SOWs so that proposal context can be assembled from related entities. | active | contextual | yes | |
| 12d622abe72dc10d… | The context_retrieve_cxu MCP tool fetches a specific CXU by ID so that verbatim clause insertion can be performed. | active | contextual | yes | |
| 147e0044e6d3b101… | The context_query MCP tool is intended to answer natural-language questions against the internal knowledge graph, such as asking for standard SLA terms. | active | contextual | yes | |
| 48385a1d5af37e39… | The roadmap backlog already contains a specification for the PYRANA Context Engine MCP Server under idea c5676045-4787-4ea9-847c-82f46d3c868f, and the internal knowledge graph is its first deployment target. | active | contextual | yes | |
| 851a072418ac71ce… | The query context sets_exclude field provides explicit exclusions from retrieval, with FINANCIAL_OPS shown as an example excluded set. | active | contextual | yes | |
| 7d22842d6c15748f… | The query context sets_include field specifies which sets to include in retrieval, exemplified by CONTRACT_STANDARDS, CLIENTS:PATRICK, and SALES_POSITIONING. | active | contextual | yes | |
| 5cd7534399e3fb5a… | The query context client_scope field is optional and narrows retrieval to a client sub-set, with PATRICK shown as an example client scope. | active | contextual | yes | |
| 5eb756e1954d5714… | The query context task_type field is used for set prioritization, with proposal_generation given as an example task value. | active | contextual | yes | |
| b34af510ac9b9492… | The query context agent_type field should identify the querying agent as cowork, claude_code, or scheduled_task so that behavior can be scoped by agent class. | active | contextual | yes | |
| 128630ab8f1af178… | The query context schema includes a user field set to an email identity, exemplified by "ankur@zeroth.technology", so that graph queries can be associated with a specific user principal. | active | contextual | yes | |
| 6331eeb2f17892ff… | The CONTRACT_STANDARDS executed set should be restricted to Leadership and Legal because executed contracts may contain negotiated terms. | active | contextual | yes | |
| 99056aad1a3b5fc9… | The CONTRACT_STANDARDS templates set should be accessible to all team members because contract templates are reusable across the team. | active | contextual | yes | |
| b41c5210a1ea6fcc… | The deployment is single-tenant for Zeroth Agents internal use, but the knowledge graph should still enforce access scoping across set categories so that information access remains controlled by sensitivity and need-to-know boundaries. | active | contextual | yes | |
| b24f7f2e88b8445b… | The number of PostgreSQL rows is estimated to equal the CXU count plus metadata rows. | active | contextual | yes | |
| b8f5697da6f347bf… | The number of Qdrant vectors is estimated to be equal to the CXU count. | active | contextual | yes | |
| 60c40713901c135a… | The Neo4j graph is estimated to contain about 100 non-CXU nodes for hierarchy, agents, and memory structures. | active | contextual | yes | |
| 3449fec1ee6f5388… | The source corpus is estimated to include more than 150 files. | active | contextual | yes | |
| 74eccab6b00428c1… | The system is estimated to have about 15 named relationship types and more than 500 relationship instances. | active | contextual | yes | |
| 15aaf8f221c81c22… | The system is estimated to contain 12 top-level sets and approximately 15 sub-sets. | active | contextual | yes | |
| 4ca32fa902911884… | After six months, the estimated total number of CXUs is between 3,000 and 5,000. | active | contextual | yes | |
| e6545a06fa237e69… | The estimated initial load for the system is between 1,200 and 2,500 total CXUs. | active | contextual | yes | |
| bfc7c97b2202b747… | The Granola integration should flag action items and decisions with high confidence after extracting meeting knowledge. | active | contextual | yes | |
| 3ad70212c7107d07… | The Granola integration should auto-assign extracted CXUs to CLIENTS and PROJECTS sets based on meeting attendees and the meeting title. | active | contextual | yes | |
| 4555d2411e586e29… | The Granola integration should extract CXUs using the zeroth-internal profile when processing meeting transcripts. | active | contextual | yes | |
| c245d3b5a582bf3b… | After each meeting, the Granola integration should query Granola for the transcript as the first step in transcript processing. | active | contextual | yes | |
| c6c0fde634a912f2… | Granola meeting integration should be configured for continuous extraction from meeting transcripts. | active | contextual | yes | |
| a0f98103cd7d87c0… | The Google Drive sync job should flag potential conflicts between new and existing CXUs for human review. | active | contextual | yes | |
| 0f3de235bce00fe9… | The Google Drive sync job should assign newly extracted CXUs to appropriate sets after processing changed documents. | active | contextual | yes | |
| d09c55a68c9cbf23… | The Google Drive sync job should run the extraction pipeline on changed documents after detecting new or modified files. | active | contextual | yes | |
| 7a5cc233bb42bbfd… | A periodic Google Drive sync job should monitor both Drive folders for new or modified files as the first step of synchronization. | active | contextual | yes | |
| 5ab6cc5e46fbf159… | Once the knowledge graph is populated, the daily research task should query the context engine directly for platform knowledge instead of reading a flat file so that platform context remains dynamic and always current. | active | contextual | yes | |
| 95c80d03afc4d5ad… | The scheduled task pyrana-daily-research-review already reads PYRANA_PLATFORM_ARCHITECTURE_REFERENCE.md as part of the daily research pipeline. | active | contextual | yes | |
| 0569d09d23496856… | The MCP tool context_search_entities is intended to find all entities related to a term such as Patrick or code interpreter. | active | contextual | yes | |
| b015099371c2d2fd… | The MCP tool context_feedback is intended to report whether a retrieved CXU was helpful or unhelpful so that Q-values can be updated. | active | contextual | yes | |
| 96c5a525cacb0e06… | The MCP tool context_publish_cxu is intended to publish a pre-formed CXU, such as one produced during contract review. | active | contextual | yes | |
| f565b255da094440… | The MCP tool context_publish is intended to publish new CXUs derived from meeting notes or newly added documents. | active | contextual | yes | |
| 210bd71ae60c48cc… | The MCP tool context_graph_traverse is intended to walk from Client to Projects to SOWs so that proposal context can be built from graph relationships. | active | contextual | yes | |
| e0a99be07f42ae20… | The MCP tool context_retrieve_cxu is intended to fetch a specific CXU by ID so that verbatim clause insertion can be performed. | active | contextual | yes | |
| 5719f3f00b2d76d9… | The MCP tool context_query is intended to answer natural-language questions against the internal knowledge graph, such as asking for standard SLA terms. | active | contextual | yes | |
| f50fbf2b843ac9bf… | The internal knowledge graph is the first deployment target for the PYRANA Context Engine MCP Server. | active | contextual | yes | |
| 84b824de56e91cba… | The roadmap backlog already contains a specification for the PYRANA Context Engine MCP Server under roadmap idea c5676045-4787-4ea9-847c-82f46d3c868f. | active | contextual | yes | |
| 942cf53f373fbf1a… | The query context schema includes an optional sets_exclude field that can explicitly exclude sets such as FINANCIAL_OPS from a knowledge graph query. | active | contextual | yes | |
| 70d0fb53cf215604… | When an agent such as Cowork, Claude Code, or a scheduled task queries the knowledge graph, it should declare a query context so that retrieval can be scoped by user, agent type, task, client, and set filters. | active | contextual | yes | |