Document explicit memory destinations and membership privacy
Finalize enterprise export documentation after PR merge
Document safe enterprise exports and validation waiver
docs: refresh wiki for current Dense-Mem behavior
Refactor telemetry and documentation for v2.6.1 synchronous Remember lifecycle - Updated Telemetry.md to clarify telemetry metrics and their meanings, focusing on synchronous Remember calls and terminal outcomes. - Revised Using-Dense-Mem.md to reflect the synchronous nature of the Remember function and its immediate return of terminal outcomes. - Enhanced Vector-Search-And-Ranking.md to specify the embedding process for v2.6.1 and its implications on search document states. - Modified Write-Pipeline.md to detail the synchronous Remember lifecycle, including the handling of submission statuses and the terminal result structure. - Clarified the public status contract and error handling in the Remember process, ensuring accurate representation of operational failures and quarantine procedures.
Refactor documentation for Remember assessor and v2.6 contract updates - Updated terminology from "normalizer" to "assessor" throughout Request-Lifecycle-And-Security.md and Technical-Reference.md. - Clarified the role of the Remember assessor in handling structured requests and responses, including grounding repair and predicate reconciliation. - Revised sections to emphasize the closed-schema validation and the rejection criteria for submissions. - Enhanced telemetry documentation to reflect changes in metrics related to Remember acknowledgements and assessment outcomes. - Adjusted examples in Using-Dense-Mem.md and Write-Pipeline.md to align with the new v2.6 contract requirements, including idempotency key usage and evidence handling. - Removed references to deprecated features and clarified the operational boundaries for submissions and assessments.
docs: document Remember intake boundary
docs: document access capability boundary for #261
docs: enhance idempotency and erasure details in request lifecycle documentation
Refactor documentation for clarity and accuracy across multiple files - Updated "Portals And API Keys" to reflect changes in correction workflows and removed outdated placement review sections. - Revised "Quick Start" to clarify upgrade processes and migration boundaries, emphasizing the importance of batch submissions. - Enhanced "Recall Feedback" to include new snapshot retention requirements and feedback submission constraints. - Adjusted "Recall Ranking" references for consistency in technical documentation. - Modified "Relationship Model And Lifecycle" to clarify submission processes and error handling, including updates to flowcharts. - Updated "Request Lifecycle And Security" to reflect changes in normalizer responses and submission handling. - Revised "Technical Reference" to clarify submission requirements and error handling, removing references to outdated processes. - Adjusted "Telemetry" metrics to align with recent changes in submission dispositions. - Updated "Using Dense Mem" to clarify the normalizer's role and submission processes, removing references to outdated states. - Revised "Write Pipeline" to reflect changes in submission handling, including the removal of the awaiting review state and updates to error handling.
Enhance documentation for private-memory features and OAuth integration - Added `MCP_PUBLIC_BASE_URL` and `PRIVATE_MEMORY_RETENTION_DAYS` to Configuration.md with detailed descriptions. - Expanded Data-Model-And-Storage.md to include private-memory lifecycle and erasure processes. - Introduced OAuth and private-erasure end-to-end testing scenarios in Development-And-Testing.md. - Updated MCP-Clients.md to clarify team-scoped and unscoped MCP URLs, OAuth protected-resource discovery, and authentication requirements. - Documented migration details and submission diagnostics in Operations-Runbook.md. - Added private-memory erasure operations and API details in Portals-And-API-Keys.md. - Clarified OAuth bearer authentication for MCP in Request-Lifecycle-And-Security.md. - Specified private-memory erasure boundaries and operations in Technical-Reference.md. - Updated User-Portal-SSO.md to include new routes for private-memory erasure and credential deletion. - Enhanced Using-Dense-Mem.md with detailed status response information for submissions.
Refactor documentation for Dense-Mem API and workflows - Updated the Request-Lifecycle-And-Security.md to clarify untrusted prompt data and added Memory Space Authorization section. - Revised Technical-Reference.md to enhance clarity on contract principles, relationship requirements, and evidence handling. - Enhanced User-Portal-SSO.md with new endpoints for credential management and clarified SSO member session behavior. - Improved Using-Dense-Mem.md to reflect changes in evidence submission and relationship proposals, emphasizing logical relationships. - Updated Write-Pipeline.md to align with new evidence and relationship handling, including changes to submission states and assessment processes.
Refactor authentication and authorization model across documentation - Updated references to "profile" to "identity" and "membership" to reflect changes in the authentication model. - Clarified the role of permanent owner aliases in team and credential management. - Removed legacy transport options and clarified the use of the official SDK transport in various documents. - Enhanced descriptions of API key and credential management in the user portal and control portal sections. - Revised the PostgreSQL schema documentation to reflect the removal of semantic references and the introduction of direct foreign keys. - Improved clarity on the handling of SSO sessions and their relationship with team memberships and credentials. - Updated operational runbook to include new migration details and job retention policies.
docs: enhance documentation for v2.5 identity model, migration history, and error handling
docs: enhance documentation on embedding reconciliation, search convergence, and contract schemas for clarity and accuracy
docs: update multiple documentation files for clarity and accuracy
docs: enhance clarity and detail in multiple documentation files
Refactor documentation across multiple files to enhance clarity and accuracy - Updated Operations-Runbook.md to clarify compatibility holds and semantic commit processes. - Revised Portals-And-API-Keys.md to detail API-key session management and cookie handling. - Adjusted PostgreSQL-Cluster-Safe-Schema.md to reflect changes in memory-pack export metadata. - Enhanced Public-HTTPS-And-Redis.md with details on cookie security and Traefik routing. - Modified Quick-Start.md to specify the absence of legacy placement and review tools. - Clarified Relationship-Model-And-Lifecycle.md regarding assessor responses and submission holds. - Updated Request-Lifecycle-And-Security.md to improve descriptions of session management and quarantine processes. - Revised Technical-Reference.md to remove deprecated memory-pack operations and clarify submission handling. - Enhanced User-Portal-SSO.md with details on API-key portal session management. - Updated Using-Dense-Mem.md to clarify status responses and quarantine handling. - Refined Write-Pipeline.md to improve explanations of semantic hold lifecycles and assessment processes.
Refine data model and security event descriptions, enhance development documentation, and clarify community assembly processes - Updated `evidence_security_events` to specify the nature of append-only scans and audit events. - Improved clarity in the Development and Testing documentation regarding injection boundaries and quarantine processes. - Enhanced the Evaluation document to specify the conditions under which unsafe corpus input is rejected. - Revised the Home documentation to better outline the intake rejection and submission processes. - Expanded the Recall Ranking document to clarify community assembly and relationship fallback mechanisms. - Updated the Relationship Model and Lifecycle documentation to reflect changes in evidence security event handling. - Clarified request lifecycle and security processes, emphasizing the role of deterministic scans and quarantine protocols. - Improved the Write Pipeline documentation to specify conditions for intake rejections and the handling of hidden controls.
Refactor code structure for improved readability and maintainability
docs: document evidence lifecycle contract
docs: define assessor correction and provider retries
Document V2.4 one-call placement architecture
docs: clarify semantic memory tool visibility and HTTP route restrictions
Document tier-free Relationship architecture
docs: consolidate upgrade guidance
Document durable cross-profile conflict review
Document V2 search contract generations
Update semantic memory v2 architecture docs