Release v2.0.10

What's Changed

• feat(memreader/LLM): add backup config for openai memreader by @fridayL in #1246
• feat(Scheduler): change Scheduler default llm to general_llm by @fridayL in #1252
• fix: The issue of the file name being stored as part of the memory as well by @whipser030 in #1222
• fix(graph_dbs): add missing status parameter to Neo4jCommunityGraphDB… by @baranchen in #1179
• feat(memcube): single cube use general_llm by @fridayL in #1253
• fix: Fix the issue where scheduled tasks are not using Redis problem by @wustzdy in #1267
• feat: add test cov ci by @CarltonXiang in #1272
• Feat: add openclaw plugin publish workflow with prebuilds by @CarltonXiang in #1278
• feat: Sync memos-local-openclaw from native_memos by @syzsunshine219 in #1279
• feat:optimzie search_by_embedding && logs by @wustzdy in #1284
• fix(openclaw): strip tags from assistant messages by @hijzy in #1287
• fix(memos-local-openclaw): normalize Windows paths when resolving better-sqlite3 by @hijzy in #1265
• fix(memos-local-openclaw): respect OPENCLAW_STATE_DIR in viewer migrate by @hijzy in #1262
• fix(memos-local-openclaw): detect Anthropic provider in OpenClaw fallback config by @hijzy in #1263
• feat: repair bug 2.0.10 by @Wang-Daoji in #1282
• fix(memos-local-openclaw): respect OPENCLAW_STATE_DIR in fallback model config by @hijzy in #1264
• fix(OpenClaw Local Plugin): resolve search and SQLite installation issues by @tangbotony in #1251
• Detect allowPromptInjection and optimize auto-recall handling by @lcpdeb in #1285
• feat(memreader): update prefermem config by @fridayL in #1294
• fix(openclaw): detect Anthropic provider and use correct API format by @zerone0x in #1260
• feat(memos-local-openclaw): This PR merges the local plugin 0317 branch (openclaw-local-plugin-20260317). by @syzsunshine219 in #1289
• Add Git Subtree of MemOS-Cloud-OpenClaw-Plugin by @XiaohuiSu in #1297
• Dev 20260319 v2.0.10 by @CaralHsi in #1255

New Contributors

• @baranchen made their first contribution in #1179
• @lcpdeb made their first contribution in #1285
• @zerone0x made their first contribution in #1260

Full Changelog: v2.0.9...v2.0.10

Release v2.0.9

What's Changed

• feat: add auto_cleanup_working; add image_id/file_info by @CaralHsi in #1169
• Fix/return message by @CaralHsi in #1175
• Feat/merge main 2.0.9 by @CaralHsi in #1190
• feat(memreader): add general configs and training memreader model by @fridayL in #1185
• feat(memreader): callback image llm parser to general llm by @fridayL in #1195
• feat:optimize export_graph by @wustzdy in #1197
• feat(memreader): change image llm in file-content as default model by @fridayL in #1198
• feat: add openclaw-local-plugin version (memos-local-openclaw) by @syzsunshine219 in #1205
• fix: project_id/manager_id issue by @CaralHsi in #1208
• feat:optimize pool && log by @wustzdy in #1210
• fix: handle string-type message in suggestion handler by @CaralHsi in #1214
• Dev 20260309 v2.0.9 by @CaralHsi in #1238

New Contributors

• @syzsunshine219 made their first contribution in #1205

Full Changelog: v2.0.8...v2.0.9

Release v2.0.8

What's Changed

• fix: downgrade AuthConfig partial-init log from WARNING to INFO by @anatolykoptev in #1052
• fix: activation memory config crashes get_default() with OpenAI by @anatolykoptev in #1051
• fix: Add Unicode sanitization for cloud embedders by @anatolykoptev in #1048
• chore: sync from main by @CaralHsi in #1147
• feat: Optimize the chunk logic of the knowledge base by @whipser030 in #1135
• feat: transfer pref to polar db by @Wang-Daoji in #1145
• feat: optimzie polardb ThreadedConnectionPool by @wustzdy in #1152
• fix: Use relativity instead of score for preference memory by @hijzy in #1153
• fix: image bug; single item in multi-mudal-reader has no embedding; by @CaralHsi in #1154
• fix: Stop throwing error when embedding is missing in add_nodes_batch… by @bittergreen in #1157
• feat: align general/naive memory with tree-memory by @CaralHsi in #1159
• feat:update logs by @wustzdy in #1161
• feat:optimize search_by_fulltext by @wustzdy in #1164
• fix: delete useless graph recall by @hijzy in #1167
• feat:optimize search_by_fulltext by @wustzdy in #1170
• feat:optimize search_by_fulltext by @wustzdy in #1171
• fix: The feedback function fails when calling the search interface due to the failure in passing the 'user_name' parameter. by @whipser030 in #1174
• feat:optimize config by @wustzdy in #1176
• feat: add persistent local memory plugin for OpenClaw (memos-local-openclaw) by @tangbotony in #1182
• docs: update READMEs and bump memos-local-openclaw-plugin to v1.0.0 by @tangbotony in #1184
• chore: change version number to v2.0.8 by @wustzdy in #1186
• Dev 20260302 v2.0.8 by @CaralHsi in #1148

Full Changelog: v2.0.7...v2.0.8

Release v2.0.7

What's Changed

• fix: add full text search for neo4j db by @hijzy in #1095
• fix: Add toggle for fulltext retrieval path by @hijzy in #1096
• fix: Add toggle for fulltext retrieval path (#1096) by @CaralHsi in #1098
• fix: add full text search for neo4j db (#1095) by @CaralHsi in #1099
• feat: add return_fields parameter to search methods (#955) by @lailoo in #1097
• feat: add optimize user_name && key words && log by @wustzdy in #1111
• hotfix: An error occurred when adding the edge of the graph. by @whipser030 in #1109
• fix: fix add/update log judgment — scope by user_name an… by @glin93 in #1116
• fix: set default relativity by @hijzy in #1119
• fix: remove unused edges & optimize logs by @wustzdy in #1123
• fix: File memory parsing to output a list-type result by @whipser030 in #1125
• feat: sync from main by @CaralHsi in #1126
• fix: use merged_from to correctly identify add/update memory logs by @glin93 in #1118
• fix: fix memory_type validation error by @bittergreen in #1130
• fix: user_name by @CaralHsi in #1131
• fix: get embedding of pref mem from db instead of recompute by @hijzy in #1133
• fix: replace bare except with except Exception in thread_safe_dict_segment by @haosenwang1018 in #1112
• docs: update README.md by @eltociear in #898
• feat: optimize get_edges by @wustzdy in #1138
• chore: change version number to 2.0.7 by @bittergreen in #1140
• Dev 20260224 v2.0.7 by @CaralHsi in #1127

New Contributors

• @lailoo made their first contribution in #1097
• @haosenwang1018 made their first contribution in #1112
• @eltociear made their first contribution in #898

Full Changelog: v2.0.6...v2.0.7

Release v2.0.6

What's Changed

• fix: allow get_user_names_by_memory_ids supports all types of db by @tangg555 in #1063
• 📃 docs: update the ReadME by @XiaohuiSu in #1070
• feat: optimization v2.0.6 by @Wang-Daoji in #1071
• feat: add keyword search by @hijzy in #1073
• fix: replace relativity with score for pref_mem by @hijzy in #1077
• fix: support cases when there is both field condition and logical operators by @endxxxx in #1076
• fix: llm call enable thinking bug by @Wang-Daoji in #1080
• fix: support hours, minutes, and seconds to the filter field. by @endxxxx in #1081
• fix: llm call enable thinking bug by @Wang-Daoji in #1082
• fix: llm call enable thinking bug by @Wang-Daoji in #1083
• fix: preference return results by @CaralHsi in #1088
• feat: optimization v2.0.6 by @Wang-Daoji in #1090
• Dev 20260209 v2.0.6 by @Wang-Daoji in #1091

Full Changelog: v2.0.5...v2.0.6

Release v2.0.5

What's Changed

• feat: Initialize data structures and class for managing memory versions by @bittergreen in #992
• fix: avoid adding fileurl to memoryvalue by @whipser030 in #995
• feat: add delete_node_by_mem_cube_id && recover_memory_by_mem_kube_id by @wustzdy in #1001
• feat: add PostgreSQL + pgvector graph database backend by @anatolykoptev in #966
• fix: The "tags" lack the key words related to the memory content. by @whipser030 in #999
• fix: add ruff check code && poetry run ruff format by @wustzdy in #1007
• fix: parallel chunk large memory items and process those chunks, but … by @endxxxx in #1011
• fix: polardb parameter bug by @wustzdy in #1013
• feat: opimize branch 2.0.5 by @Wang-Daoji in #1016
• fix: The "total nodes" information returned by the "get memory" interface is inaccurate. by @whipser030 in #1018
• docs: add MemOS Cloud OpenClaw Plugin to Resources by @XiaohuiSu in #1023
• feat: add neo4j and neo4j_community function by @wustzdy in #1028
• feat: Add correlation threshold filtering and maximum limit for retrieved memories by @hijzy in #1006
• feat: skill memory by @Wang-Daoji in #1031
• refactor: (scheduler) modularize handlers and search pipelines by @fancyboi999 in #1004
• fix: set default relativity by @hijzy in #1036
• fix: The "total nodes" information returned by the "get memory" interface is inaccurate. by @whipser030 in #1037
• feat: change code struct && update neo4j by @wustzdy in #1040
• feat: modify sdk by @anpulin in #1034
• fix: format error by @CarltonXiang in #1044
• hotfix: Code restoration, with the original content of the knowledge base replaced by the sources. by @whipser030 in #1055
• feat: skill memory by @Wang-Daoji in #1042
• fix: format error by @CarltonXiang in #1046
• feat: skill memory by @Wang-Daoji in #1056
• feat: mem read rawfile handling by @glin93 in #1058
• feat: Recall the original text of the knowledge base content rather than retrieving the generated memory. by @tangg555 in #1057
• add MemOS Cloud OpenClaw Plugin into the ReadME by @XiaohuiSu in #1059
• feat: skill memory by @Wang-Daoji in #1062
• feat: opimize branch 2.0.5 by @Wang-Daoji in #1064
• feat: skill memory by @Wang-Daoji in #1065
• feat: fix RawFileMemory bug by @Wang-Daoji in #1066
• feat: merge v2.0.5 to main by @CarltonXiang in #1061
• chore: update version to 2.0.5 by @CarltonXiang in #1067

New Contributors

• @anatolykoptev made their first contribution in #966
• @XiaohuiSu made their first contribution in #1023
• @anpulin made their first contribution in #1034

Full Changelog: v2.0.4...v2.0.5

Release v2.0.4

What's Changed

• feat: add openwork by @CaralHsi in #918
• merge dev v2.0.3 into main by @hijzy in #947
• feat: add recaller for fast-add process by @bittergreen in #958
• chore: update pr template by @CarltonXiang in #959
• feat: skill memory by @Wang-Daoji in #964
• fix: playground chat bug by @CaralHsi in #968
• feat: fix get_nodes by @wustzdy in #971
• feat: skill memory by @Wang-Daoji in #972
• feat: skill memory by @Wang-Daoji in #974
• fix: playground chat bug by @CarltonXiang in #975
• fix: playground chat bug & search_by_embedding in add process by @CaralHsi in #973
• feat: Add NLI model for fast detection of duplication/conflict by @bittergreen in #970
• feat: skill memory by @Wang-Daoji in #981
• feat: add mmr deduplication by @hijzy in #954
• feat: skill memory by @Wang-Daoji in #982
• fix: embedding service timeout occasionally by @endxxxx in #983
• feat: skill memory by @Wang-Daoji in #984
• feat: adjust skill prompt and fix some bugs by @CaralHsi in #988
• fix: skill thread bug by @CaralHsi in #990
• Dev 20260126 v2.0.4 by @CarltonXiang in #976

New Contributors

• @hijzy made their first contribution in #947

Full Changelog: v2.0.3...v2.0.4

Release v2.0.3

What's Changed

• feat: refactor & reorganize examples with unified structure and updated demos by @CaralHsi in #903
• fix: refine config for rerank input is too long by @whipser030 in #920
• fix: modifying api examples by @CaralHsi in #917
• fix: component_init initialize scheduler every time by @tangg555 in #915
• fix: parse in sim-struct chat-data; unify image analysis prompt with context injection & improve LLM robustness by @CaralHsi in #913
• fix: json loads error by @whipser030 in #937
• feat: sync main by @CaralHsi in #944
• feat: model error traceback by @CarltonXiang in #942
• feat: delete_time && delete_record_id for neo4j,polardb,neo4j_community by @wustzdy in #945

Full Changelog: v2.0.2...v2.0.3

Release v2.0.2

What's Changed

• fix: knowledge base adopt raw text by @whipser030 in #836
• Feat/optimize cloud service api by @Wang-Daoji in #839
• Feat/fix palyground bug by @Wang-Daoji in #841
• refactor & fix bugs: fix a range of bugs in scheduler and revise fine add apis by @tangg555 in #840
• feat: merge dev 0112 by @CarltonXiang in #854
• feat: delete result by @CarltonXiang in #852
• fix: unuse rerank by @whipser030 in #855
• fix: backtrack knowledge retrieval by @whipser030 in #857
• Feat/tool mem related by @Wang-Daoji in #864
• docs: fix server start cmd by @Nyakult in #858
• fix: Qdrant empty when using neo4j-community by @OhhhhPi in #843
• fix: use knowledge embedding score rerank by @whipser030 in #867
• fix: optimization_and_fix by @Wang-Daoji in #868
• Dev 20260112 v2.0.2 by @CarltonXiang in #877
• feat: merge dev into main; add scheduler examples, API enhancements, and retrieval dedup by @CaralHsi in #775
• chore: update version number to v2.0.2 by @CarltonXiang in #879
• feat: prevent repeated memories via archival and activated-only retrieval by @CaralHsi in #880
• feat: optimization and fix branch by @Wang-Daoji in #882
• feat: LLM/plugin retrieval bug, prevent redundant memories via archival with activated-only retrieval, and update implicit preference logic by @endxxxx in #893

New Contributors

• @OhhhhPi made their first contribution in #843

Release v2.0.1

What's Changed

• update requirements by @pursues in #772
• update scheduler and add operation for dehallucination by @tangg555 in #769
• fix: update README.md by @zZhangSir in #774
• feat： update readme by @zZhangSir in #776
• feat: add export_graph data page by @wustzdy in #778
• fix: optimize Neo4j Community Edition support and enhance MCP environment loading by @fancyboi999 in #754
• fix: add feedback change to preference by @whipser030 in #771
• fix: improve chat playground stability and chat handler initialization by @Wang-Daoji in #770
• fix: update README.md by @zZhangSir in #786
• update: requirement,Dockerfile by @pursues in #784
• Feat: add OpenAI log by @CarltonXiang in #785
• Feat/dedup playground display by @Wang-Daoji in #789
• hotfix: redis dependency in scheduler by @tangg555 in #781
• add get_user_names_by_memory_ids api by @Wang-Daoji in #790
• feat: add batch delete by @wustzdy in #787
• feat: add timer log by @CarltonXiang in #793
• feat: add dedup search param by @glin93 in #788
• Dev zdy 1226 page by @wustzdy in #796
• requirements add neo4j by @pursues in #792
• Main zhq readme by @zZhangSir in #794
• update docker.compose.yml by @pursues in #798
• feat: fix requirements by @lijicode in #800
• Patch: get_memory adds the page size parameter function and the filtering of user id by @whipser030 in #801
• Dev zdy 1229 by @wustzdy in #802
• feat: add get_user_names_by_memory_ids for polardb && neo4j by @wustzdy in #803
• feat: add export_graph total by @wustzdy in #804
• add: get_memory return edges and count of items by @whipser030 in #805
• Scheduler: address some issues to run old scheduler example and kv cache example by @tangg555 in #797
• fix: issues caused by no reading default use_redis from env by @tangg555 in #799
• add: milvus return data pagination by @whipser030 in #806
• feat: update source return and chunk settings by @fridayL in #808
• add neo4j by @pursues in #809
• Scheduler: update exampels by @tangg555 in #807
• feat: add delete_node_by_prams filter by @wustzdy in #810
• fix: merge all preference by @whipser030 in #811
• feat: update _build_filter_conditions_sql in conditions && build_cypher_filter_condition filter by @wustzdy in #812
• feat: _build_filter_conditions_sql filter by @wustzdy in #813
• fix: update deprecated APIs for chonkie v1.4.0 and qdrant-client v1.16.0 by @zhixiangxue in #705
• feat: update code format by @fridayL in #814
• Feat/optimize cloud service api by @Wang-Daoji in #816
• fix: [PrefEval Evaluation] propagate --lib and --version arguments in search and response modes by @fancyboi999 in #780
• fix: fix context error and empty embedding by @fridayL in #817
• Feat/optimize cloud service api by @Wang-Daoji in #818
• fix: logs context and empty embedding by @fridayL in #819
• Feat/optimize cloud service api by @Wang-Daoji in #820
• add exist_user_name for neo4j.py by @wustzdy in #821
• Feat/optimize cloud service api by @Wang-Daoji in #822
• fix reranker by @pursues in #823
• Feat/optimize cloud service api by @Wang-Daoji in #825
• feat: add filter time query by @wustzdy in #826
• add getMemory sdk by @wustzdy in #827
• feat: Merge from main (some hot-fix) by @CaralHsi in #832
• feat: add OpenAI token log by @CarltonXiang in #831
• fix: feedback llm output fail to load json by @whipser030 in #833
• fix: Use env exchange overrides for all scheduler messages by @glin93 in #834
• feat: support single-assistant mem-reader by @CaralHsi in #835
• feat: tmp optimize by @Wang-Daoji in #846
• Dev 20260109 v2.0.1 by @CarltonXiang in #844
• feat: modify readme by @zhouhanshu in #838
• chore: bump version to v2.0.1 by @CaralHsi in #849

New Contributors

• @zhixiangxue made their first contribution in #705
• @zhouhanshu made their first contribution in #838

Full Changelog: v2.0.0...v2.0.1