| b1535fc2f77e382e… | The ongoing ingestion phase must continuously route new Granola meeting notes into the CLIENTS and PROJECTS sets so that current client and project knowledge remains updated over time. | active | contextual | yes | |
| 46d939b89335b7b2… | When an agent such as Cowork, Claude Code, or a scheduled task queries the knowledge graph, it should declare its context so that query scoping can be applied appropriately. | active | contextual | yes | |
| ee0fc34735eb9683… | IP_PATENTS should be restricted to Leadership and Engineering because intellectual property strategy is confidential. | active | contextual | yes | |
| 7b564ef5167d56d8… | FINANCIAL_OPS should be restricted to Leadership only because financial data is sensitive. | active | contextual | yes | |
| a5b5dd2620b98c19… | CLIENTS:* data should be accessible only to assigned team members because client information is governed by need-to-know access. | active | contextual | yes | |
| 2caaeef36d8927de… | COMPANY_POLICIES should be accessible to all team members because policies apply to everyone in the organization. | active | contextual | yes | |
| 0728ffdce2ed382d… | Executed CONTRACT_STANDARDS documents should be restricted to Leadership and Legal because executed contracts may contain negotiated terms. | active | contextual | yes | |
| 09023371fa759c62… | CONTRACT_STANDARDS templates should be accessible to all team members because templates are reusable across the organization. | active | contextual | yes | |
| 571da25ec24cda7e… | The PYRANA_PLATFORM set should be accessible to all team members because platform knowledge is shared across the organization. | active | contextual | yes | |
| b6795f14fe04f506… | The deployment is single-tenant for Zeroth Agents internal use only, but the knowledge graph should still enforce access scoping across content categories. | active | contextual | yes | |
| 81d28d23b35c3fbe… | The AmbientAgent should enable gap detection to identify missing knowledge areas such as absent client security assessments or missing pricing models for the Code Interpreter tier. | active | contextual | yes | |
| 406a1c66dcb0be2c… | MSA terms and SLA commitments are an example of CXUs that are often co-retrieved for proposal generation, illustrating the retrieval patterns the LearningAgent should learn. | active | contextual | yes | |
| 47b72d86f426337b… | The LearningAgent should enable TEMPR memory to learn which CXUs are most frequently retrieved together so that co-retrieval patterns can improve proposal generation workflows. | active | contextual | yes | |
| c3280cc05b287ab7… | The RetrievalAgent should be enabled with cross-encoder reranking so that contract clause retrieval is more precise. | active | contextual | yes | |
| dba5ea8d4fefa566… | Market estimates and projections should initialize CONTEXT_FOR edges with a Q-value of 0.45 because they are inherently uncertain sources. | active | contextual | yes | |
| 8b8dce3f193a5e7f… | Meeting transcripts should initialize CONTEXT_FOR edges with a Q-value of 0.55 because they are raw records that may contain misinformation. | active | contextual | yes | |
| cfd5ce48a94f7d8f… | Draft documents should initialize CONTEXT_FOR edges with a Q-value of 0.60 because they may contain unvalidated content. | active | contextual | yes | |
| f2d77738295bd13d… | Final-version presentations should initialize CONTEXT_FOR edges with a Q-value of 0.80 because they have been reviewed and presented. | active | contextual | yes | |
| 1baeaad6bc340109… | Contract templates should initialize CONTEXT_FOR edges with a Q-value of 0.85 because they are approved documents that have not yet been executed. | active | contextual | yes | |
| 086c346dc982393f… | Signed contracts such as MSAs and SOWs should initialize CONTEXT_FOR edges with a Q-value of 0.90 because they are executed legal documents. | active | contextual | yes | |
| eb454dff4152fe02… | GitHub CLAUDE.md files should initialize CONTEXT_FOR edges with a Q-value of 0.95 because they are considered authoritative, maintained documentation. | active | contextual | yes | |
| ee6f30ae4a661c4d… | Initial Q-values for CONTEXT_FOR edges should be set according to source reliability so that retrieval confidence reflects how authoritative each source type is. | active | contextual | yes | |
| 83c78305b96935af… | Text2Cypher is intended to handle open-ended questions such as lessons learned from pharma engagements and changes in positioning over the past year. | active | contextual | yes | |
| e6b1434fc5f006db… | Strategy 3, Text2Cypher, should start with lower weight and increase as the graph matures so that open-ended queries can be supported more heavily when graph coverage improves. | active | contextual | yes | |
| 4bdc76b211ae7e07… | Edge-filter Cypher should answer which SOWs reference the code interpreter by filtering IMPLEMENTS edges to PYRANA_PLATFORM:CODE_INTERPRETER so that implementation-linked statements of work can be found. | active | contextual | yes | |
| 52effd3530eb0e93… | Edge-filter Cypher should answer project queries for life sciences clients by filtering from CLIENTS:LIFE_SCIENCES through the SCOPED_BY relationship so that client-sector project sets can be enumerated. | active | contextual | yes | |
| 267c40db679c3241… | Strategy 2, Edge-filter Cypher, should receive medium weight because it is useful for relationship-heavy queries that depend on filtering graph edges. | active | contextual | yes | |
| 3c7f477236351b6d… | Template Cypher should support COMPANY_POLICIES lookups for questions about security policy topics so that policy answers can be retrieved from a known template pattern. | active | contextual | yes | |
| 775e3adc0867dd2d… | Template Cypher should support traversal of IMPLEMENTS relationships for questions about platform components used by Patrick so that component usage can be answered through known graph patterns. | active | contextual | yes | |
| 3f11f654eda7263c… | Template Cypher should support a template for CONTRACT_STANDARDS requirements when users ask about standard payment terms so that standard contractual requirements can be retrieved directly. | active | contextual | yes | |
| 106fb1822da6cc2d… | Strategy 1, Template Cypher, should receive high weight because most internal queries are expected to match known patterns, enabling reliable template-based retrieval. | active | contextual | yes | |
| 58a966e555593642… | The retrieval pipeline should use a three-strategy configuration tuned for internal knowledge patterns so that different query types are handled by the most suitable retrieval method. | active | contextual | yes | |
| 1512e5b2a54e2810… | Daily research pipeline findings should be ingested into the PYRANA_PLATFORM set as new research insights so that platform knowledge incorporates ongoing research outputs. | active | contextual | yes | |
| a9093df8d5505805… | The system should automatically classify and extract new documents added to the shared Drive so that newly introduced content enters the knowledge graph without manual triage. | active | contextual | yes | |
| 9cebb9a24adac816… | The ongoing ingestion phase includes continuously routing new meeting notes from Granola into the CLIENTS and PROJECTS sets so that current client and project knowledge remains updated over time. | active | contextual | yes | |
| f190a59e3ba4109b… | A Template Cypher retrieval pattern should answer questions about security policy topics by using a COMPANY_POLICIES lookup template so that policy queries map to a predefined retrieval pattern. | active | contextual | yes | |
| 7783cf3f226d097a… | A Template Cypher retrieval pattern should answer questions about platform components used by Patrick by traversing the IMPLEMENTS relationship so that client-platform linkage queries follow a known graph path. | active | contextual | yes | |
| f08a2b9fa4505a97… | A Template Cypher retrieval pattern should answer questions about standard payment terms by using a CONTRACT_STANDARDS requirements template so that contractual requirement queries map to a known retrieval path. | active | contextual | yes | |
| 7823ace8121bdb39… | Strategy 1, Template Cypher, should have high weight because most internal queries are expected to match known patterns so that template-based retrieval is prioritized. | active | contextual | yes | |
| d3ab4735f26b9f00… | The retrieval system should configure a three-strategy retrieval pipeline for internal knowledge patterns so that query handling aligns with expected internal use cases. | active | contextual | yes | |
| 672ec8c3902e0cc4… | Within Phase 4, daily research pipeline findings must be ingested into PYRANA_PLATFORM as new research insights so that platform knowledge is continuously enriched. | active | contextual | yes | |
| 3b51c6567a4e4869… | Within Phase 4, new documents added to Drive must be classified and extracted automatically so that newly added content enters the graph without manual routing. | active | contextual | yes | |
| 09b0b1122dce0d76… | Within Phase 4, new Granola meeting notes must be continuously ingested into CLIENTS and PROJECTS so that ongoing client and project discussions are captured without interruption. | active | contextual | yes | |
| 46e0968d7f783171… | Phase 4 of the extraction priority order must support ongoing ingestion for new meeting notes, new Drive documents, and daily research findings so that the knowledge graph stays continuously updated. | active | contextual | yes | |
| 02fd76aa9351dafb… | Within Phase 3, remaining shared-drive content must be extracted into various sets after the named operational sources so that uncategorized residual content is still ingested. | active | contextual | yes | |
| 63cf402f0a48ef2a… | Within Phase 3, patent documents must be extracted into IP_PATENTS so that intellectual property knowledge is incorporated before remaining shared-drive content. | active | contextual | yes | |
| d23feb5407b47b04… | Within Phase 3, financial documents must be extracted into FINANCIAL_OPS so that financial operational knowledge is loaded during the operational phase. | active | contextual | yes | |
| 86b263c99e98e400… | Within Phase 3, security and compliance documents must be extracted into SECURITY_COMPLIANCE so that compliance-related knowledge is ingested as an operational priority. | active | contextual | yes | |
| 8da6535ec77af432… | Within Phase 3, remaining sales materials must be extracted into SALES_POSITIONING so that non-core sales knowledge is added during the operational stage. | active | contextual | yes | |
| e754c9410a788790… | Phase 3 of the extraction priority order must ingest operational knowledge during weeks 5 through 6, covering sales, security, financial, patent, and remaining shared-drive content. | active | contextual | yes | |