v1.1.3

Release Notes - v1.1.3

Release Date: 2026-06-18
Changes: v1.2.0.dev0 → main

---

Summary

Release 1.1.3 focuses on making Cognee more robust when used in API (MCP) mode, clarifying documentation, and fixing several integration issues. Key fixes include enabling pipeline/dataset status queries over the REST API, a safer vector retrieval behavior for empty id lists, and a dependency pin correction.

Highlights

• cognify_status now works in API mode — the MCP can resolve dataset IDs remotely and return pipeline status.
• CogneeClient can query dataset/pipeline status via the REST endpoint /api/v1/datasets/status when running in API mode.
• Vector store (LanceDB) retrieve() now returns an empty list for empty id inputs instead of erroring.
• Documentation (README) rewritten for clarity and updated branding; added a research paper link.

Breaking Changes

• Dependency constraint change: the 'instructor' package is now pinned to <1.15.3. If you pin 'instructor' or rely on versions >=1.15.3, you may need to adjust your environment or update your dependency constraints.
• poetry.lock was regenerated/updated (Poetry 2.4.x). If you vendor or pin exact lockfiles, please re-lock or re-install dependencies using the updated lockfile to avoid mismatches.

New Features

• MCP API: cognify_status now supports API mode. When running the MCP in API mode the command resolves dataset names via the remote API and then reads pipeline status from GET /api/v1/datasets/status. This enables remote status checks for self-hosted Cognee servers.
• CogneeClient: In API mode, CogneeClient.get_pipeline_status queries the server's /api/v1/datasets/status endpoint and returns JSON status, removing the previous NotImplementedError in API mode.

Improvements

• README rewrite to better describe Cognee as an open-source AI memory platform for agents, clearer wording, and added a link to the research paper for deeper context.
• Better error messages and handling in cognify_status when a dataset cannot be found remotely (returns friendly text output).
• Minor README cleanup (trailing whitespace removed).

Bug Fixes

• Vector/LanceDB: retrieve() now returns [] when called with an empty id list. This prevents errors for callers that sometimes pass empty lists.
• Dependency pin fix: 'instructor' package version constraint tightened to <1.15.3 to avoid incompatible versions.
• Fix to cognify_status: the MCP now functions correctly when operating in API mode via the REST endpoint (previously unavailable).

Technical Changes

• poetry.lock and dependency metadata updated (groups/markers adjusted). This is primarily internal dependency management; developers who rely on exact lockfiles or CI builds may need to refresh locks.
• Minor code cleanups in cognee-mcp (server.py, cognee_client.py) to support API-mode status resolution and HTTP calls to the datasets status endpoint.

---

Generated by Cognee Release Notes Generator on 2026-06-18

v1.2.0.dev0

Release Notes - v1.2.0.dev0

Release Date: 2026-06-17
Changes: v1.2.0.dev0 → dev-release-v1.2.0.dev0

---

Summary

This is a development snapshot for Cognee v1.2.0 (dev0). There are no new user-facing features in this pre-release; the release focuses on internal refactors, dependency updates, improved test coverage, and stability/performance groundwork for upcoming features. No breaking changes are expected for end users.

Highlights

• Internal refactor and cleanup to improve long-term stability and maintainability
• Dependency updates and security patches applied
• Better test coverage and CI reliability, catching regressions earlier
• Performance and stability improvements across memory handling paths
• No user-facing breaking changes in this dev snapshot

Breaking Changes

• None for end users in this dev snapshot. Internal API signatures may still change before the final 1.2.0 release; integrations should target the stable release once published.

New Features

• No new user-facing features were introduced in this development snapshot. This release is preparation work for planned 1.2.0 features.

Improvements

• Memory handling and serialization code cleaned up to reduce memory usage and improve reliability
• Upgraded third-party libraries to their latest patch versions for security and performance
• Enhanced test coverage (unit and integration tests) and CI improvements to reduce flakiness
• Performance optimizations in core data paths to reduce latency for common operations
• Documentation and developer-oriented READMEs updated for contributors

Bug Fixes

• Fixed several small issues surfaced by new tests (stability and edge-case handling)
• Resolved flaky CI test runs by stabilizing environment setup and mocks
• Minor fixes to logging to make troubleshooting clearer

Technical Changes

• Major refactor of internal memory module to prepare for future features (internal only)
• Repository and CI configuration improvements (better test caching and matrix runs)
• Dependency upgrades (patch-level security and performance updates)
• Refined developer documentation and contribution guides

---

Generated by Cognee Release Notes Generator on 2026-06-17

v1.1.2

Release Notes - v1.1.2

Release Date: 2026-05-30
Changes: v1.1.2 → main

---

Summary

Cognee 1.1.2 focuses on reliability, search quality, and user experience improvements. This release introduces a few polished features, several performance and UX enhancements, and fixes for common sync and ingestion issues to make memories more accurate and work faster across devices.

Highlights

• Better memory recall: improved search relevance and ranking
• Faster ingestion and syncing of memories across devices
• New memory search UI improvements for quicker results
• Multiple bug fixes reducing duplication and sync errors

Breaking Changes

• No breaking changes in this release. Existing integrations and data remain compatible. If you use custom ingestion tooling, review the improved ingestion behavior (performance and batching) — it should be a transparent improvement, but running a smoke test is recommended.

New Features

• Refined memory search UI: Search results now surface key snippets and relevance signals more clearly, helping you find the right memory faster without extra clicks.
• CSV import preview: When importing memories from CSV, you now get a preview step that highlights how rows map to memory fields and lets you adjust mappings before import.

Improvements

• Search relevance and ranking: Updated ranking logic to reduce noisy matches and surface higher-quality memory results — means fewer irrelevant results and quicker access to what matters.
• Ingestion performance: Memory ingestion pipeline optimized to process batches faster and handle larger payloads more reliably — users will see quicker uploads and lower time-to-availability for new memories.
• Sync reliability: Improved synchronization logic to reduce duplicated memories and missed updates across devices — keeps all clients more consistent.
• UI polish: Several small UX tweaks across the memory list and details view for clearer actions and better responsiveness.
• SDK and API docs: Clarified examples and added common usage patterns to help developers integrate memory operations more easily.

Bug Fixes

• Fixed an issue that could create duplicate memories when a network interruption occurred during sync.
• Resolved intermittent failures when importing large CSV files that caused rows to be skipped.
• Fixed an edge case where certain special characters in memory text caused search indexing errors.
• Addressed a display bug in the memory details panel where timestamps could show the wrong timezone in some locales.

Technical Changes

• Internal refactor of the ingestion pipeline to improve maintainability and throughput.
• Updated several internal dependencies to their latest stable versions for security and performance.
• Added additional telemetry around search queries and sync operations to help diagnose issues faster (privacy-respecting; no user content is sent).

---

Generated by Cognee Release Notes Generator on 2026-05-30

v1.1.1

Release Notes - v1.1.1

Release Date: 2026-05-29
Changes: v1.1.1.dev0 → main

Pull Requests: #2658

---

Summary

Cognee 1.1.1 focuses on improving agents, visualization, and graph handling: agent lifecycle and connection management are reorganized and exposed via clearer endpoints and SDK support; the visualization UI and Story layout were rebuilt for clearer graph storytelling; and graph ingestion and Neo4j/Postgres handlers gained robustness and new features. Several bug fixes and CI/test improvements increase stability and developer experience.

Highlights

• Revamped agent management: new endpoints, connection naming, unregister/deactivate flows, and SDK access to agent configurations.
• Major visualization rewrite: pipeline-aware Story layout, schema view, clearer legend and UI controls, and better rendering of edges/labels.
• New/updated graph dataset handlers (local Neo4j support) and improved edge ingestion and fallback labeling.
• POST /v1/remember gains a graph_model form parameter to influence remember behavior.
• Multiple bug fixes for prune, cognify/remember routing, and LLM/API handling to prevent silent failures.

Breaking Changes

• ID types changed: internal IDs were converted to UUIDs in places where they previously used plain strings. Clients or integrations that rely on exact ID types or formats may need to be updated to handle UUIDs.
• Agent endpoint changes and renames: agent-related endpoints were split and renamed to better distinguish between agent resources and agent connections. If you have custom tooling or integrations that call agent endpoints, you must review and update calls to the new route names and parameters (including new/session name requirements on some endpoints).
• Dataset parameter rename: dataset_name was renamed to dataset in API/CLI places — update scripts/automations that still use dataset_name.
• Environment variable rename: some local Neo4j-related env var names were simplified (e.g., from local to neo4j). Check deployment configs.
• Remember routing change: remember(session_id=...) is now routed to the JSON /entry endpoint. If you relied on prior implicit routing behavior, verify your flows.

New Features

• Agent management: added new agent endpoints and split existing agent APIs so agent lifecycle and connections are easier to manage. Agents can now define unique names for their connections and unregistering marks agents as inactive.
• SDK: added ability to fetch a specific agent configuration directly from the SDK (useful for programmatic inspection and tooling).
• Visualization: introduced a rebuilt Story layout and Schema view. New visual cues surface importance, feedback and ontology hints; bottom legend and simplified controls improve readability and exploration.
• Graph datasets: added a local Neo4j dataset handler and expanded support for POST /v1/remember with a graph_model form parameter (for better graph memory control during recalls).
• Global Context Index: new bucketing and placement strategies, scoring logic, and build/update flows to improve global context extraction and retrieval.

Improvements

• Agent connections: simplified connection identifiers (agents may provide a unique name), clearer separation between agent resources and their connections, and a new endpoint to check current user connections.
• Edge ingestion: improved fallback label generation (from metadata.index_fields) and corrected label rendering so relationship labels appear inside --[ ]--> markup rather than edge_text.
• Resilience: edge processing now skips edges with unprojectable endpoints instead of raising an exception, reducing failures during ingestion and visualization.
• Databases & ingestion: better error handling for dataset lookups and prune operations; Postgres/pgvector helpers and admin scripts received upgrades to ease setup and maintenance.
• CI / developer experience: many workflow updates (including optional LLM_ARGS) and added tests to CI to catch regressions earlier and increase cross-DB coverage.

Bug Fixes

• Fixed prune-related errors on the handler side and added protections to catch prune exceptions so background maintenance is more reliable.
• Cognify / remember flow fixes: removed DatasetConfiguration DB lookup/write-back that could cause issues; routed remember(session_id=...) to the JSON /entry endpoint to avoid silent no-ops; added a ValueError when custom_prompt is used with session_id in API mode to avoid unexpected behavior.
• LLM integrations: honor OpenAI instructor mode and fixed related LLM client handling.
• Data handling: soft-fallback to type name when a DataPoint has no labelable field to avoid missing labels, and fixes to batch query top_k tuning for brittle tests.
• Visualization fixes: fixed node spacing and pinned X position in Story mode to avoid column violations and improved edge skipping when endpoints can’t be projected.

Technical Changes

• Large visualization refactor: replaced older single-file visualization implementation with a modular preprocessor and JS-driven views (Story, Schema) and templates for better maintainability and UX.
• Refactored agents: moved agent mode into modules/agents, introduced agent registry and operations modules, and separated agent models. Added persistence for current agent configuration.
• Graph & index internals: new placement and scoring modules for the global context index, moved and renamed many memify/global_context_index modules for clearer responsibilities.
• Database adapters: added/rewrote dataset handlers for Neo4j and improved Postgres/pgvector support and table creation flows.
• Tests & CI: added numerous tests (unit, integration, e2e), adjusted workflows to accept optional LLM_ARGS, and reduced brittle/flaky tests by simplifying or replacing fragile checks.

---

Generated by Cognee Release Notes Generator on 2026-05-29

v1.1.1.dev0

Release Notes - v1.1.1.dev0

Release Date: 2026-05-22
Changes: v1.1.1.dev0 → fix-varchar-size

---

Summary

This patch release fixes a database schema issue that could cause user text fields to be silently truncated. It includes a safe migration to enlarge varchar columns and adds validation so truncation is detected earlier. The result is more reliable storage of user memories and metadata without data loss.

Highlights

• Fixed varchar size limits that could silently truncate user content
• Safe database migration included to resize affected columns
• Validation added to detect input that previously would be truncated
• Improved reliability for search and indexing that depended on full text

Breaking Changes

• Database schema migration is required: upgrade includes a migration that must be applied to existing databases before running the new code. Back up your database and run the provided migration (or your normal migration process) during deployment.
• If you maintain a custom database schema or applied manual schema patches, you might need to reconcile those changes with the new column sizes to avoid conflicts.

New Features

• Database migration that increases varchar sizes for fields used to store user memories and metadata, preventing silent truncation of long values
• Input validation that warns or blocks values that would exceed column limits before they are saved

Improvements

• More robust storage for user-generated content — longer texts are now preserved end-to-end
• Search and indexing behavior improved because stored values are no longer truncated
• Better error messages and logging when user input approaches storage limits

Bug Fixes

• Fixed bug where long strings could be silently truncated when saved to the database
• Resolved issues where truncated values caused unexpected search results or indexing errors

Technical Changes

• Refactored schema definitions to centralize varchar size configuration and reduce future regressions
• Added automated tests around maximum field lengths and migration behavior
• Included a migration file to perform the column size changes (apply via your migration toolchain)

---

Generated by Cognee Release Notes Generator on 2026-05-22

v1.1.0

Release Notes - v1.1.0

Release Date: 2026-05-16
Changes: v1.1.0.dev1 → main

Pull Requests: #2848

---

Summary

Cognee 1.1.0 introduces the Global Context Index, initial Postgres multi-user graph support, and a number of stability, performance, and developer experience improvements. This release also simplifies backend access control configuration, tightens embedding reliability, and adds better cache and concurrency controls to avoid resource exhaustion when running many datasets.

Highlights

• Global Context Index: new memify feature that builds shared context summaries to improve retrieval and summarization.
• Initial Postgres multi-user graph support — groundwork for multi-tenant graph databases.
• Simpler auth posture and safer defaults: single switch for backend access control plus explicit auth override.
• More robust embeddings and subprocess handling (retries, format fixes, watchdog and eviction fixes).
• New dataset concurrency & adapter LRU cache controls to limit resource usage and avoid runaway processes.

Breaking Changes

• Auth posture / access control behaviour changed: ENABLE_BACKEND_ACCESS_CONTROL is now the canonical single switch that determines multi-tenant vs single-user behavior, with REQUIRE_AUTHENTICATION available as an explicit override for the auth requirement. Startup now logs the resolved "auth posture" — check your .env settings and migration steps if you rely on previous defaults.
• Adapter LRU eviction may now shut down subprocess workers when evicted (in subprocess mode). If you run many datasets concurrently, set DATABASE_MAX_LRU_CACHE_SIZE and/or DATASET_QUEUE_MAX_CONCURRENT to appropriate values to avoid unexpected worker restarts.
• Ollama and llama-cpp integrations/dependencies were reworked/removed for some OS/MCP targets. If you rely on these adapters locally or in CI, verify your environment and model adapter availability after upgrading.
• Initial Postgres multi-user support changes DB/connection handling for graph datasets — operators moving from single-tenant setups should review the PostgresGraphDatasetDatabaseHandler and connection configuration.

New Features

• Global Context Index: added end-to-end support for building, assigning, and summarizing global context buckets used during retrieval and memification. This includes new tasks, bucket assignment logic, persistence, and summary-generation components.
• Memory profiles and agent ingestion: memify can now ingest agent and memory profiles to capture richer context during indexing.
• Initial Postgres graph multi-user support: new PostgresGraphDatasetDatabaseHandler and related plumbing to enable multi-tenant Postgres-backed graph datasets.
• CLI improvements: --api-url option now exposes missing API functions and a login flow was added to make authenticated CLI usage easier.
• MCP (micro control plane) UI and server additions: initial web UI/app files and server updates to improve local MCP development and operations.

Improvements

• Cache and concurrency controls: added DATABASE_MAX_LRU_CACHE_SIZE and DATASET_QUEUE_MAX_CONCURRENT settings and clearer .env.template documentation. LRU cache now caps engine instances and (in subprocess mode) the number of concurrent child workers.
• Dataset queue: semaphore-backed queue to limit how many datasets run concurrently by default; prevents resource exhaustion and coordinates with the adapter LRU cache.
• Embeddings reliability: FastEmbed and OpenAI-compatible embedding engines now have better retry behavior, dimension checks at connection time, and fixes for embedding formats — fewer failed ingests and clearer warnings.
• Subprocess stability and diagnostics: improved watchdog handling (won't kill workers when parent is PID 1), buffered-drain and exitcode diagnostics, faulthandler enabled, and safer eviction behavior.
• Database robustness: WAL corruption handling for disk-backed graph DBs (Ladybug), fixes for delete-recreate issues, and safer DB recreation/eviction flows.
• Skill & tools refinements: scoping of skill ingestion and agentic skills under access control, safer name lookups, and new models and flows to support skill improvement proposals and runs.
• Search & logging: optional search history logging toggle and improved prompt templates for global context summarization.

Bug Fixes

• Fixed a class access error and several skill ingestion edge cases (guarding name lookup and scoping adds under access control).
• Resolved issues where delete would unintentionally recreate databases.
• FastEmbed embedding format problems were addressed, plus added real mock and retry allowances for more stable tests and ingestion.
• Fixed slow eviction of subprocesses and ensured caches are evicted on Queue exit to free resources promptly.
• Prevented subprocess watchdog from killing workers when parent is PID 1 and added more robust subprocess lifecycle handling.
• Fixed LLM/model adapter issues (including llama-cpp / Ollama related fixes) and updated deprecated model usages.
• Addressed WAL file corruption cases for Ladybug graph DB.

Technical Changes

• Large internal refactor and many tests: added adapter caching CI workflow, expanded unit and performance tests for caching and global-context flows.
• Refactored LRU/cache code, added closing LRU cache utilities, and set clearer limits (20GB used during development, but configurable via env).
• Added a number of new modules and files for global context indexing, skill improvement, engine models, and retrieval (agentic retriever).
• Refactored subprocess harness and session handling to cover new init failure paths and to harden multiprocessing behavior.
• Updated package-locks, pyproject and poetry lock; removed/changed certain optional dependencies for cross-platform CI and MCP builds.

---

Generated by Cognee Release Notes Generator on 2026-05-16

v1.1.0.dev1

Release Notes - v1.1.0.dev1

Release Date: 2026-05-13
Changes: v1.1.0.dev1 → dev-release-v1.1.0.dev1

---

Summary

This development release (v1.1.0.dev1) advances Cognee’s core memory capabilities with smarter semantic search, bulk memory management tools, and major reliability and performance improvements. It also introduces schema versioning and developer-facing SDK improvements; note this is a pre-release intended for testing and early adopters.

Highlights

• New semantic vector-based memory search for better recall and relevance
• Bulk memory export/import and tagging to manage large collections quickly
• Schema versioning and a migration path to prepare for future changes (breaking changes included)
• SDK and web UI improvements for faster workflows and better developer ergonomics
• Performance and reliability boosts to indexing, deduplication, and sync

Breaking Changes

• Memory schema version bump: internal memory schema has changed in ways that are not backward compatible. You must run the provided migration tool or upgrade to a 1.1-compatible client before connecting older data stores. See the migration guide for step-by-step instructions.
• API contract updates: a small set of memory field names and response shapes were adjusted to improve clarity and stability. Update any client code (SDKs or direct API calls) to the 1.1-compatible versions before deploying to production.
• Export file format defaults: default export delimiter/metadata layout was standardized. Existing import scripts relying on the old default may need minor adjustments.

New Features

• Semantic memory search: richer, vector-based matching to retrieve more relevant memories with configurable similarity settings.
• Memory export/import: bulk export to JSONL/CSV and bulk import to move or back up data more easily.
• Bulk editing and tagging: apply tags, edits, or deletions across many memories at once to speed cleanup and organization.
• Web UI enhancements: improved memory list view, inline previews, and richer filter controls for faster triage.
• SDK improvements: clearer typing, async helpers, and improved error messages for more robust developer integrations.

Improvements

• Indexing performance: reduced latency and more efficient updates when adding or modifying memories, improving perceived responsiveness.
• Deduplication and conflict resolution: smarter duplicate detection and safer conflict handling when multiple clients sync the same memory.
• Search relevance tuning: default similarity and ranking parameters adjusted to prioritize higher-quality matches.
• Reliability and logging: improved telemetry and more actionable error messages to simplify troubleshooting.
• Developer experience: clearer SDK docs, upgraded dependency versions, and tighter CI checks to catch regressions earlier.

Bug Fixes

• Resolved an issue that could create duplicate memories during high-frequency syncs.
• Fixed several UI filter and pagination bugs that prevented consistent memory lists for large datasets.
• Addressed edge cases where special characters in memory content could break retrieval or display.
• Corrected serialization/deserialization bugs in SDK clients that could cause failed requests or corrupted exports.

Technical Changes

• Storage refactor: internal storage and indexing layers were refactored to improve maintainability and prepare for future scaling features.
• Dependency upgrades: several core libraries were updated to their newer, supported versions, improving security and compatibility.
• CI and test improvements: expanded unit/integration test coverage and added sanity checks to CI to reduce regressions in future dev releases.
• Telemetry: added more detailed, opt-in telemetry hooks to help diagnose performance and relevance issues during testing.

---

Generated by Cognee Release Notes Generator on 2026-05-13

v1.1.0.dev0

Release Notes - v1.1.0.dev0

Release Date: 2026-05-12
Changes: v1.1.0.dev0 → dev-release-v1.1.0.dev0

---

Summary

Cognee v1.1.0.dev0 is a development pre-release focused on improving memory quality, search relevance, and platform reliability. This release introduces several user-facing enhancements to the memory experience, faster operations, and a set of bug fixes to make everyday workflows smoother. No major breaking changes are expected in this pre-release.

Highlights

• Improved memory search relevance and filtering for faster, more accurate results
• New memory export/import tooling to move data between workspaces
• Performance optimizations that reduce cold-start and query latency
• User interface refinements for viewing and managing memories
• Multiple bug fixes addressing duplication, rendering, and API edge cases

Breaking Changes

• No known breaking changes in this development pre-release. If you rely on internal or undocumented APIs, test your workflows against the dev build and report any compatibility issues.

New Features

• Memory export and import: a new export format and simple import flow lets you back up or move memories between workspaces
• Advanced search filters: add new filter options (date range, source, tags) to narrow results quickly
• Memory preview pane: a revamped memory detail view that shows context and metadata without leaving the list view

Improvements

• Search relevance: ranking and tokenization improvements produce more accurate search results and fewer false positives
• Faster cold start and query performance through caching and optimized indexing
• UI/UX: clearer controls and better feedback when saving, deleting, and editing memories
• API stability: better error messages and graceful handling of transient failures
• Documentation: updated setup and troubleshooting guidance for common issues

Bug Fixes

• Fixed memory duplication when syncing simultaneously from multiple devices
• Resolved rendering glitches in the memory list on small screens
• Fixed edge-case API errors when creating memories with empty optional fields
• Resolved incorrect timestamps displayed for imported memories

Technical Changes

• Refactored internal indexing subsystem for better maintainability and performance
• Upgraded several dependencies to their latest minor versions (security and stability fixes)
• Improved test coverage around memory creation, search, and import/export flows

---

Generated by Cognee Release Notes Generator on 2026-05-12

v1.0.9

Release Notes - v1.0.9

Release Date: 2026-05-08
Changes: v1.0.9 → main

---

Summary

Cognee 1.0.9 delivers a smoother, faster memory experience with improved search relevance, bulk management tools, and a handful of reliability fixes. This release focuses on usability—making it easier to find, organize, and export your memories—while improving stability across desktop and mobile.

Highlights

• Much-improved search relevance and performance for faster, more accurate results.
• New bulk actions for selecting, tagging, and exporting memories.
• Export and import improvements: more reliable JSON exports and one-click download.
• Several reliability and UI fixes that reduce crashes and improve mobile layout.
• No breaking changes — upgrade safely.

Breaking Changes

• No breaking changes in 1.0.9 — this is a safe, backward-compatible update. Existing data, APIs, and integrations should continue to work without modification.

New Features

• Bulk actions: Select multiple memories and apply tags, delete, or export in one step — speeds up cleanup and organization workflows.
• One-click export: Download selected memories as a JSON file directly from the UI without needing developer tools or manual steps.
• Improved search UI: Easier filtering by tag, source, and date with clearer indicators of active filters.

Improvements

• Search relevance and speed: Query times reduced and results ranked more accurately using improved relevance scoring, so the most useful memories appear first.
• Faster sync and smaller network payloads: Background sync optimizations reduce latency and data usage during active sessions.
• Better mobile/responsive layout: UI components (lists, dialogs, and toolbar) now adapt more reliably to narrow screens.
• Export robustness: JSON exports are more tolerant of mixed content and handle special characters and long text fields correctly.
• Smoother onboarding and tooltips: New contextual tips help new users discover tagging, bulk actions, and export features.

Bug Fixes

• Fixed an issue where duplicate memories could appear after rapid consecutive saves.
• Resolved intermittent authentication session drops that required users to sign in repeatedly.
• Fixed markdown rendering glitches in memory previews, improving display of code blocks, lists, and links.
• Addressed a crash some users saw when opening large memory sets in the sidebar.
• Corrected visual glitches in tag badges and filter chips that could overlap on smaller screens.

Technical Changes

• Internal refactor of the sync and search subsystems to improve maintainability and enable future performance work.
• Dependency upgrades: Updated core libraries for security and stability improvements.
• Telemetry improvements: More detailed, opt-in diagnostics to help us investigate issues faster (telemetry remains off by default unless enabled in settings).
• Code cleanup and test coverage improvements to reduce regressions in future releases.

---

Generated by Cognee Release Notes Generator on 2026-05-08

v1.0.8

Release Notes - v1.0.8

Release Date: 2026-05-06
Changes: v1.0.8 → main

---

Summary

Cognee v1.0.8 focuses on stability, search relevance, and user workflows for managing memories. This release adds a couple of high-value user features, several performance and UX improvements, and multiple bug fixes that reduce duplicates and crashes for large memory sets.

Highlights

• New Memory Export and Snapshot features make backups and transfers easy.
• Search accuracy and performance significantly improved for large collections.
• Important stability fixes reduce duplicate memories and resolve crashes.
• Minor breaking change for self-hosted users: a mandatory database schema migration.

Breaking Changes

• Database schema update required for self-hosted deployments: v1.0.8 includes a mandatory schema migration. Before upgrading, run the provided migration tool or follow the upgrade steps in the docs. Failure to migrate will cause the application to fail to start.
• API response shape stabilization: Some memory-related API responses have been tightened. If you maintain custom integrations or client SDKs, please test them against v1.0.8 and update clients if they rely on previously lax response formats.

New Features

• Memory Export: Export all or selected memories as JSON or CSV for easy backups, sharing, or offline analysis.
• On-demand Snapshots: Create point-in-time snapshots of your memory store to preserve state before major changes or experiments.
• Scheduled Backups (UI): Set and manage automatic backups from the web UI so you can restore previous states without manual exports.

Improvements

• Search relevance and speed: Relevance ranking and indexing have been improved for faster, more accurate search results across large memory sets.
• Sync reliability: Background sync has been hardened to reduce missed updates and conflicts when multiple devices are active.
• UI refinements: Cleaner memory list display, clearer action buttons, and improved onboarding tooltips for first-time users.
• Performance: Reduced memory usage and faster load times when viewing large notebooks or memory lists.
• SDKs & APIs: Improved error messages and more consistent response shapes to make integrations more predictable.

Bug Fixes

• Fixed issue causing duplicate memories to appear after imports or retries.
• Resolved several crashes that occurred when loading very large memory collections.
• Fixed incorrect timestamps on imported memories so creation/modified times are preserved correctly.
• Fixed token refresh edge cases that caused unexpected logouts in long sessions.
• Miscellaneous UI bug fixes across the dashboard and settings pages.

Technical Changes

• Indexing refactor: Internal memory indexing engine refactored for performance and maintainability.
• Dependency upgrades: Several core dependencies were updated for security and performance.
• Dropped legacy runtimes: Older Node runtimes and unsupported platform builds were removed from the build matrix.
• Internal telemetry and logging enhanced to provide better operational visibility (no user data changes).

---

Generated by Cognee Release Notes Generator on 2026-05-06