fuli-context 0.23.0 → 0.25.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +169 -979
- package/README.zh-CN.md +157 -830
- package/dist/web/assets/{AboutPage-S8QFfzpA.js → AboutPage-Dx07_Z8u.js} +1 -1
- package/dist/web/assets/{AboutSectionHeading-B5CzH4Zs.js → AboutSectionHeading-BdhpM609.js} +1 -1
- package/dist/web/assets/AgentAppsPage-5VW0uCTz.js +1 -0
- package/dist/web/assets/{AgentAppsPage-BuJWNNdL.css → AgentAppsPage-BCxrjrsE.css} +1 -1
- package/dist/web/assets/AgentAvatar-DbsG1_BY.css +1 -0
- package/dist/web/assets/AgentAvatar-DlAQP7gm.js +1 -0
- package/dist/web/assets/{AgentHand-BGGdVdd9.js → AgentHand-FPFg9-Y1.js} +1 -1
- package/dist/web/assets/AgentManagement-CLFw9oV9.css +1 -0
- package/dist/web/assets/AgentManagement-UTe4I3WA.js +7 -0
- package/dist/web/assets/AgentMentionCopy-B9eIIr3L.css +1 -0
- package/dist/web/assets/AgentMentionCopy-C5ExlutT.js +1 -0
- package/dist/web/assets/{AgentName-CUjBNQ8j.js → AgentName-C8mYbtKs.js} +1 -1
- package/dist/web/assets/AgentProfilePage-BLeyT3By.js +5 -0
- package/dist/web/assets/{AgentProfilePage-D5QUzw5N.css → AgentProfilePage-DumsVHP1.css} +1 -1
- package/dist/web/assets/AgentServicesPage-B2mDUgYZ.css +1 -0
- package/dist/web/assets/AgentServicesPage-BXi152Xc.js +1 -0
- package/dist/web/assets/{AppViewerPage-CcOlwxck.js → AppViewerPage-BR8ttEGa.js} +1 -1
- package/dist/web/assets/{BackToTop-C3KYEoCW.js → BackToTop-5_AQRMZl.js} +1 -1
- package/dist/web/assets/{BoardCardFieldsMenu-BlBWT1j7.js → BoardCardFieldsMenu-Bb_Z9a4Q.js} +1 -1
- package/dist/web/assets/{BoardMoveMenu-Bc7xC1Uh.js → BoardMoveMenu-B-f8O6cC.js} +1 -1
- package/dist/web/assets/{Breadcrumb-C_xMiD-Z.js → Breadcrumb-DtM5lZy5.js} +1 -1
- package/dist/web/assets/{CaptureChangePage-Bekp1aWg.js → CaptureChangePage-BzwcfacI.js} +1 -1
- package/dist/web/assets/ConnectionsPage-C_vgqwF2.js +1 -0
- package/dist/web/assets/ConnectionsPage-DS2PEG6K.css +1 -0
- package/dist/web/assets/{CopyFixPrompt-Bp-UX4Hq.js → CopyFixPrompt-DlarhfPK.js} +1 -1
- package/dist/web/assets/DataFormatPage-Ccy_e9rf.css +1 -0
- package/dist/web/assets/DataFormatPage-DwL4Ofc9.js +2 -0
- package/dist/web/assets/DebugTurnPage-UpK_3ZWY.js +1 -0
- package/dist/web/assets/DebugTurnsPage-C0aZg5-3.js +1 -0
- package/dist/web/assets/{DialogTitle-Lj5s6nwG.js → DialogTitle-BQXOe8EK.js} +1 -1
- package/dist/web/assets/{EmployeeRedirectPage-DnvwJeCN.js → EmployeeRedirectPage-BjWpXDSM.js} +1 -1
- package/dist/web/assets/EmployeeScopeDialog-Bt3ZlDyZ.js +1 -0
- package/dist/web/assets/EmployeeScopeDialog-U_-eNH0Z.css +1 -0
- package/dist/web/assets/{EmployeeWorkbench-BhReI8l6.css → EmployeeWorkbench-Bp9pUkfS.css} +1 -1
- package/dist/web/assets/EmployeeWorkbench-GtJIUD5G.js +4 -0
- package/dist/web/assets/FieldRoot-DRf4wM5x.js +1 -0
- package/dist/web/assets/FlaGrowth-DgVLB8hp.css +1 -0
- package/dist/web/assets/FlaGrowth-OQtTLTng.js +1 -0
- package/dist/web/assets/FlaOccupationKit-BvrKsosv.css +1 -0
- package/dist/web/assets/FlaOccupationKit-C-x4VrKk.js +1 -0
- package/dist/web/assets/{FlaTaskSheets-CNdaBy1P.js → FlaTaskSheets-CoaF5CbQ.js} +1 -1
- package/dist/web/assets/FlasPage-CgIVOGFM.js +1 -0
- package/dist/web/assets/FlasPage-H99IZ9M6.css +1 -0
- package/dist/web/assets/{GrowthLoading-Cr8xZwwO.js → GrowthLoading-VuVo2vLf.js} +1 -1
- package/dist/web/assets/{HowtoPage-Cv1YQy1s.js → HowtoPage-CZ3mOdLc.js} +1 -1
- package/dist/web/assets/{HowtoPage-t12tDDpS.css → HowtoPage-CrvMmT6K.css} +1 -1
- package/dist/web/assets/InboxPage-1ue8vA7H.css +1 -0
- package/dist/web/assets/InboxPage-zNUnnYeJ.js +6 -0
- package/dist/web/assets/{KnowledgeDeleteButton-FEyEvhlR.js → KnowledgeDeleteButton-BlUnSG3b.js} +1 -1
- package/dist/web/assets/{KnowledgeHardDeleteSection-CVwTeMEJ.js → KnowledgeHardDeleteSection-CznXBtHY.js} +1 -1
- package/dist/web/assets/KnowledgeItemEditPage-DGRNkAYd.js +1 -0
- package/dist/web/assets/{KnowledgeItemSummaryCard-49EADY_Z.js → KnowledgeItemSummaryCard-Co5JKBHl.js} +1 -1
- package/dist/web/assets/KnowledgeOrganizerPage-BWvF1GEu.js +1 -0
- package/dist/web/assets/KnowledgeOrganizerPage-CJaFkfS7.css +1 -0
- package/dist/web/assets/KnowledgePage--Z4_kAhq.js +1 -0
- package/dist/web/assets/KnowledgePeek-BgzQA9SI.js +1 -0
- package/dist/web/assets/{KnowledgePeek-Bm5jRVA9.css → KnowledgePeek-CwZVuSkf.css} +1 -1
- package/dist/web/assets/{KnowledgeTurnReceiptPage-BOz7IpaK.js → KnowledgeTurnReceiptPage-Csyjimeg.js} +1 -1
- package/dist/web/assets/KnowledgeWorkspace-0yOpWtEB.js +4 -0
- package/dist/web/assets/{KnowledgeWorkspace-DCoVFyf0.css → KnowledgeWorkspace-BgWo6lBj.css} +1 -1
- package/dist/web/assets/{MeAvatar-CYVcRiuW.js → MeAvatar-aY_q8cKs.js} +1 -1
- package/dist/web/assets/{MeLink-BdqkzhT6.js → MeLink-B0pNwVLa.js} +1 -1
- package/dist/web/assets/MenuSeparator-BR-uLovR.js +1 -0
- package/dist/web/assets/ModelSettingsPage-B9PS1Kay.css +1 -0
- package/dist/web/assets/ModelSettingsPage-BL6MsF7C.js +1 -0
- package/dist/web/assets/{NotFoundPage-Dmn32VWo.js → NotFoundPage-Cr1MaPuk.js} +1 -1
- package/dist/web/assets/NoticeRow-BXZWn4EZ.js +1 -0
- package/dist/web/assets/NoticeRow-ktVO_9Es.css +1 -0
- package/dist/web/assets/OccupationsPage-CkVqZ2pz.js +1 -0
- package/dist/web/assets/OccupationsPage-Dup7WzNh.css +1 -0
- package/dist/web/assets/OverviewPage-BkJ17OYv.js +1 -0
- package/dist/web/assets/OverviewPage-Dr4jrXW4.css +1 -0
- package/dist/web/assets/{PageColumns-DpfaPeu7.js → PageColumns-DvehZlYm.js} +1 -1
- package/dist/web/assets/PageOutline-CFGtHmOK.js +1 -0
- package/dist/web/assets/PageOutline-DG5ZYtA8.css +1 -0
- package/dist/web/assets/{PersonalProfilePage-DdBfHnqq.css → PersonalProfilePage-DnjNX2xK.css} +1 -1
- package/dist/web/assets/PersonalProfilePage-TxsNr44t.js +1 -0
- package/dist/web/assets/PersonalProjectsPage-AhgP6CL8.js +1 -0
- package/dist/web/assets/{PinIcon-B7dvh7ld.js → PinIcon-BTQsnxRY.js} +1 -1
- package/dist/web/assets/PopperContent-DtMksk76.js +1 -0
- package/dist/web/assets/{Presence-D3HKb5Hb.js → Presence-B0zMpCF4.js} +1 -1
- package/dist/web/assets/ProfileAuthorChip-BqT9Vv65.js +1 -0
- package/dist/web/assets/ProfileAuthorChip-CAWNGoW3.css +1 -0
- package/dist/web/assets/{ProjectAgentsPage-RpfiV2Fm.css → ProjectAgentsPage-CjeXFzHr.css} +1 -1
- package/dist/web/assets/ProjectAgentsPage-O-cMavcz.js +5 -0
- package/dist/web/assets/{ProjectJudgmentPolicy-GA33aSRi.js → ProjectJudgmentPolicy-CyHAu0nm.js} +2 -2
- package/dist/web/assets/{PublicProjectsPage-PJlBI8IE.js → PublicProjectsPage-Ck1QSR0R.js} +1 -1
- package/dist/web/assets/RecentNotices-vJobVOcm.js +1 -0
- package/dist/web/assets/{RoundtablesPage-DT5DUXsa.js → RoundtablesPage-C3dYekQ5.js} +6 -6
- package/dist/web/assets/{RoundtablesPage-DK9A9J-y.css → RoundtablesPage-DL-IOngN.css} +1 -1
- package/dist/web/assets/RovingFocusGroup-CUqyjFSs.js +1 -0
- package/dist/web/assets/{RovingFocusItem-CFd7tXzD.js → RovingFocusItem-CyyuWXjE.js} +1 -1
- package/dist/web/assets/{SessionMigrationPage-BZ1lEDQr.js → SessionMigrationPage-DukDM0TP.js} +1 -1
- package/dist/web/assets/SessionRecordsPage-DZ5ZBiuS.js +1 -0
- package/dist/web/assets/SettingsPage-CWGk0XZ_.js +1 -0
- package/dist/web/assets/SettingsPage-Cj3pEWcd.css +1 -0
- package/dist/web/assets/SharedJefaReportPage-DUY03VLS.js +1 -0
- package/dist/web/assets/TagVocabularyPage-BKLFJiSu.js +1 -0
- package/dist/web/assets/TaskBlockerClear-BBmPbomO.css +1 -0
- package/dist/web/assets/TaskBlockerClear-BxOiiuK8.js +1 -0
- package/dist/web/assets/TaskBlockerNotice-iUiD3BuZ.js +1 -0
- package/dist/web/assets/{TaskReason-LQ0CFiKu.js → TaskReason-D5j7L8tl.js} +1 -1
- package/dist/web/assets/TeamCard-CDunbIrP.js +1 -0
- package/dist/web/assets/TeamCard-CihQsZZ1.css +1 -0
- package/dist/web/assets/{TeamCardsWidget-DGX9k0S6.js → TeamCardsWidget-CUW-fPn7.js} +1 -1
- package/dist/web/assets/TeamIsland-CmpDmIX2.js +9 -0
- package/dist/web/assets/{TeamIsland-BBQbpozz.css → TeamIsland-DvtcsAsb.css} +1 -1
- package/dist/web/assets/{Teleport-C1upABpY.js → Teleport-Dzdr8umQ.js} +1 -1
- package/dist/web/assets/{TextField-ZnmRqG0h.js → TextField-BPhEpEv9.js} +1 -1
- package/dist/web/assets/TonborgDelegationSwitch-C1nkJqkJ.js +1 -0
- package/dist/web/assets/TonborgDelegationSwitch-DeqERWoo.css +1 -0
- package/dist/web/assets/{UiChip-CizeOu1Q.js → UiChip-BEQ067LV.js} +1 -1
- package/dist/web/assets/{UiDialog-CGTzxn3-.js → UiDialog-JjqzOA0z.js} +1 -1
- package/dist/web/assets/{UiDisclosure-ueMA4jl0.js → UiDisclosure-L6GEJHSw.js} +1 -1
- package/dist/web/assets/{UiIcon-sD1hUsOF.js → UiIcon-vgD4v8q8.js} +1 -1
- package/dist/web/assets/UiIconButton-BzyMc14P.js +1 -0
- package/dist/web/assets/{UiSegmented-BZ-jqnPE.js → UiSegmented-CIIm3KoV.js} +1 -1
- package/dist/web/assets/{UiWeight-BRO3oVMZ.js → UiWeight-D-2iZ_Kc.js} +1 -1
- package/dist/web/assets/VirtualDirectoryList-BJecsQct.css +1 -0
- package/dist/web/assets/VirtualDirectoryList-BV8KsA-Z.js +1 -0
- package/dist/web/assets/{VisuallyHidden-B8wfJ_mU.js → VisuallyHidden-BxKxuXRF.js} +1 -1
- package/dist/web/assets/VisuallyHiddenInput-DD4vhlqT.js +1 -0
- package/dist/web/assets/WidgetSlot-Dw5nWlb7.js +2 -0
- package/dist/web/assets/WorkbenchListHead-B0TjHsTO.js +1 -0
- package/dist/web/assets/WorkbenchListPage-BX6Z9gkB.css +1 -0
- package/dist/web/assets/WorkbenchListPage-D0xX-YNl.js +1 -0
- package/dist/web/assets/_plugin-vue_export-helper-B4pFJCwG.js +9 -0
- package/dist/web/assets/agent-apps-B1Je-mpY.js +1 -0
- package/dist/web/assets/agent-occupation-CjfQqSzd.js +1 -0
- package/dist/web/assets/{agent-pins-8KX0ynzy.js → agent-pins-DxHlJ96z.js} +1 -1
- package/dist/web/assets/agent-profile-BNPLhDog.css +1 -0
- package/dist/web/assets/{agent-roster-records-PfAgndHp.js → agent-roster-records-Bi5HWDif.js} +1 -1
- package/dist/web/assets/agent-work-summary-o_9-QWBL.js +1 -0
- package/dist/web/assets/{agent-workflows-DQlqVOoA.js → agent-workflows-DgdbJqpg.js} +1 -1
- package/dist/web/assets/{attach-closest-edge-D2qlayPy.js → attach-closest-edge-DUc_VUVO.js} +1 -1
- package/dist/web/assets/{attention-store-BnZ8l3kk.js → attention-store-Bod20yM3.js} +1 -1
- package/dist/web/assets/{badge-DX_xrLV6.js → badge-tMy2Qr94.js} +1 -1
- package/dist/web/assets/{button-CEkPBC1E.js → button-DnKFsImW.js} +1 -1
- package/dist/web/assets/{card-DNQ0_6vT.js → card-Ccu7HCNb.js} +1 -1
- package/dist/web/assets/{catalog-KnpBkv5t.js → catalog-CJUV8OJ6.js} +1 -1
- package/dist/web/assets/checkbox-D1R6kTUB.js +1 -0
- package/dist/web/assets/{client-CSQcWFRh.js → client-D-G2i5VU.js} +2 -2
- package/dist/web/assets/client-links-store-UgmraFBZ.js +1 -0
- package/dist/web/assets/combobox-C7QLYGzs.js +1 -0
- package/dist/web/assets/console-DgRc8Itw.js +1 -0
- package/dist/web/assets/{conversation-launcher-settings-CXFE-q_A.js → conversation-launcher-settings-CZAfuu8B.js} +1 -1
- package/dist/web/assets/copy-mention-C0EJhDPy.js +1 -0
- package/dist/web/assets/{dialog-CxhKMo0D.js → dialog-Bbq93W68.js} +1 -1
- package/dist/web/assets/{dist-B9GqMC5C.js → dist-Dd8yFhv6.js} +1 -1
- package/dist/web/assets/{dist-DIJqkL4O.js → dist-PpsW5ZE3.js} +1 -1
- package/dist/web/assets/dropdown-menu-1fGUL0V3.js +1 -0
- package/dist/web/assets/{field-LNsBZqam.js → field-DjVv_wIS.js} +1 -1
- package/dist/web/assets/{format-bytes-Bqefdt37.js → format-bytes-BKq_E0Hs.js} +1 -1
- package/dist/web/assets/hover-card-BO-929hq.js +1 -0
- package/dist/web/assets/{identity-DRfhWjty.js → identity-DIG-uOiU.js} +1 -1
- package/dist/web/assets/index-DF0TcfPY.css +1 -0
- package/dist/web/assets/index-RWEx1M91.js +24 -0
- package/dist/web/assets/{input-UoZ-0Xne.js → input-DmwRmYO2.js} +1 -1
- package/dist/web/assets/is-equal-8VOV02sH.js +1 -0
- package/dist/web/assets/{label-DpjbTOxt.js → label-BvwPUTS4.js} +1 -1
- package/dist/web/assets/label-table-CDzYRZpc.js +1 -0
- package/dist/web/assets/{model-CpuIG8Du.js → model-0TZRbkN2.js} +1 -1
- package/dist/web/assets/model-BF_ZMPLi.js +1 -0
- package/dist/web/assets/model-BjUERpoQ.js +3 -0
- package/dist/web/assets/model-CaR7VPg7.js +1 -0
- package/dist/web/assets/{model-pJ6k3zYD.js → model-Dfe81jZT.js} +1 -1
- package/dist/web/assets/{model-BYKHVgJr.js → model-DpkZUZUL.js} +1 -1
- package/dist/web/assets/{model-13BQndfF.js → model-Tmn6HvL-.js} +1 -1
- package/dist/web/assets/{model-B5g0CFcM.js → model-y7OUvtKz.js} +1 -1
- package/dist/web/assets/motion-C1fZrqmI.js +1 -0
- package/dist/web/assets/{motion-C-ZimCUS.js → motion-mAmbRj_M.js} +1 -1
- package/dist/web/assets/{motion-preference-BQiiwhab.js → motion-preference-Dz9S1bZa.js} +1 -1
- package/dist/web/assets/npm-copies-BLRc0FXW.js +1 -0
- package/dist/web/assets/{client-connection-model-DVwA-GIu.css → npm-copies-GRyMAduA.css} +1 -1
- package/dist/web/assets/occupation-catalog-_SnxxRbK.js +1 -0
- package/dist/web/assets/occupation-kits-DFC6Auuc.js +1 -0
- package/dist/web/assets/{option-CcHPBOo6.js → option-DC8frdbr.js} +1 -1
- package/dist/web/assets/{orchestration-filters-DmF4Qs8E.js → orchestration-filters-UeiAHJqn.js} +1 -1
- package/dist/web/assets/overview-data-Cq1gefY5.js +1 -0
- package/dist/web/assets/page-refresh-DdRJaxMv.js +1 -0
- package/dist/web/assets/{page-title-NXphgeQI.js → page-title-Cm9u_qw1.js} +1 -1
- package/dist/web/assets/person-profile-BH2AzOfw.js +1 -0
- package/dist/web/assets/{popover-xSRzcSmY.js → popover-DW__YCSE.js} +1 -1
- package/dist/web/assets/popup-host-ctKYu9Ze.js +1 -0
- package/dist/web/assets/profile-model-DtISOJ-X.js +1 -0
- package/dist/web/assets/{project-board-layout-DBQeP_QP.js → project-board-layout-CTM8kAgq.js} +1 -1
- package/dist/web/assets/{project-capture-BhYsDWwJ.js → project-capture-B3pNhpFF.js} +1 -1
- package/dist/web/assets/{project-origin-D8jy8EF_.js → project-origin-BO8Ngccf.js} +1 -1
- package/dist/web/assets/quick-edit-api-B169r5Qb.js +1 -0
- package/dist/web/assets/radio-group-DqA46cXQ.js +1 -0
- package/dist/web/assets/review-api-CJaNgPMA.js +1 -0
- package/dist/web/assets/review-list-BbV-Mle7.js +1 -0
- package/dist/web/assets/review-store-DSQpb5P8.js +1 -0
- package/dist/web/assets/{role-names-PECaMfQo.js → role-names-By0xdfTn.js} +1 -1
- package/dist/web/assets/roundtable-live-u9qUU6aA.js +1 -0
- package/dist/web/assets/{runtime-dom.esm-bundler-f39i44Mt.js → runtime-dom.esm-bundler-DhFy4Djs.js} +1 -1
- package/dist/web/assets/save-file-oGyIokMI.js +1 -0
- package/dist/web/assets/select-TqPBladU.js +1 -0
- package/dist/web/assets/{session-launch-offers-X9b0DwF1.js → session-launch-offers-CtQeY4Qk.js} +1 -1
- package/dist/web/assets/{settings-restart-BnwcTv3i.js → settings-autosave-C7vW3_SP.js} +1 -1
- package/dist/web/assets/settings-restart-p12HUUve.js +1 -0
- package/dist/web/assets/{shared-report-BxNVLalR.js → shared-report-DQBSigHB.js} +3 -3
- package/dist/web/assets/{source-adapters-aupsL146.js → source-adapters-DPJzumzC.js} +1 -1
- package/dist/web/assets/{switch-B3Y3oE_n.js → switch-D7w_0h3m.js} +1 -1
- package/dist/web/assets/tabs-DR8k0PN9.js +1 -0
- package/dist/web/assets/{task-card-fields-DAx6zMLz.js → task-card-fields-X1ZYCm_j.js} +1 -1
- package/dist/web/assets/{task-labels-BdDFvSAy.js → task-labels-LuXoQhF9.js} +1 -1
- package/dist/web/assets/{task-reason-7WnPreht.js → task-reason-ChpbGtyx.js} +1 -1
- package/dist/web/assets/{team-sheets-Cr0ZTPw3.js → team-sheets-DYCQqa8I.js} +1 -1
- package/dist/web/assets/{textarea-CRzFqL9a.js → textarea-D53VYanK.js} +1 -1
- package/dist/web/assets/ui-D4nAhcmu.css +1 -0
- package/dist/web/assets/ui-DmNbL8MW.js +1 -0
- package/dist/web/assets/{useAgentRecords-CMtpd3Fj.js → useAgentRecords-Hy9nqMs7.js} +1 -1
- package/dist/web/assets/useAgentWorkSummary-mloD4p12.js +1 -0
- package/dist/web/assets/{useApi-CROJJdhE-CRCgNOqq.js → useApi-CROJJdhE-DnUGUb8O.js} +1 -1
- package/dist/web/assets/{useClipboardCopy-DkqKK3wY.js → useClipboardCopy-BTuQMeWI.js} +1 -1
- package/dist/web/assets/{useFocusReturn-BZpVmQnM.js → useFocusReturn-4hUmSzj9.js} +1 -1
- package/dist/web/assets/{useForwardExpose-Di-MnbTz.js → useForwardExpose-BDmW7q7u.js} +1 -1
- package/dist/web/assets/{useForwardPropsEmits-L9XcB5DG.js → useForwardPropsEmits-BzT2luD4.js} +1 -1
- package/dist/web/assets/useForwardScopeId-DZCnSswO.js +1 -0
- package/dist/web/assets/useKnowledgeConfirm-L8wnJCuX.js +1 -0
- package/dist/web/assets/useKnowledgeItemPeek-C_huXt6E.js +1 -0
- package/dist/web/assets/{useMinimumLoadingDisplay-CWPiOaDl.js → useMinimumLoadingDisplay-Dkl8NQru.js} +1 -1
- package/dist/web/assets/usePageTools-CbDOdIrg.js +1 -0
- package/dist/web/assets/usePrimitiveElement-tPQM7XJs.js +1 -0
- package/dist/web/assets/{useProjectAgentRoster-CWVpSsQ3.js → useProjectAgentRoster-BBmRvcgr.js} +1 -1
- package/dist/web/assets/{useRoutePeek-DxLVksbJ.js → useRoutePeek-CRbCDqsm.js} +1 -1
- package/dist/web/assets/useTeamPeople-D_5cxxf1.js +1 -0
- package/dist/web/assets/{useTeamWork-BTZd_wKs.js → useTeamWork-CXEQwF4n.js} +1 -1
- package/dist/web/assets/{useTypeahead-pEZNpBls.js → useTypeahead-CVYuxPOD.js} +1 -1
- package/dist/web/assets/{utils-BsxOjD1Q.js → utils-BnrWDlfm.js} +1 -1
- package/dist/web/assets/utils-BsGjKHV8.js +1 -0
- package/dist/web/assets/verification-DbCo_dnY.js +1 -0
- package/dist/web/assets/{version-db3QWWjt.js → version-Ci3l8Vnb.js} +1 -1
- package/dist/web/assets/{vue-router-B79HWW1l.js → vue-router-D4_ms6om.js} +1 -1
- package/dist/web/assets/workbench-card-Dt0820yW.css +1 -0
- package/dist/web/index.html +90 -74
- package/docs/agent-conversations-and-collaboration.md +55 -58
- package/docs/agent-interface-architecture.md +7 -2
- package/docs/agent-roundtable.md +9 -8
- package/docs/client-connections.md +85 -8
- package/docs/client-links.md +16 -1
- package/docs/employee-agents.md +1 -1
- package/docs/judgment-and-autonomy.md +2 -0
- package/docs/public-personal-architecture.md +1 -1
- package/graph-provider/fuli_graph/app.py +2 -0
- package/graph-provider/fuli_graph/graph_field_contract.py +6 -3
- package/graph-provider/fuli_graph/graph_format_migrations.py +6 -0
- package/graph-provider/fuli_graph/graph_models.py +11 -62
- package/graph-provider/fuli_graph/graph_query.py +46 -105
- package/graph-provider/fuli_graph/knowledge_batch_confirmation.py +7 -5
- package/graph-provider/fuli_graph/knowledge_duplicate_models.py +24 -0
- package/graph-provider/fuli_graph/knowledge_duplicate_rules.py +157 -0
- package/graph-provider/fuli_graph/knowledge_duplicates.py +357 -0
- package/graph-provider/fuli_graph/knowledge_item_routes.py +3 -1
- package/graph-provider/fuli_graph/knowledge_item_view.py +45 -2
- package/graph-provider/fuli_graph/knowledge_item_view_models.py +21 -0
- package/graph-provider/fuli_graph/knowledge_management.py +6 -1
- package/graph-provider/fuli_graph/knowledge_outdated.py +15 -5
- package/graph-provider/fuli_graph/knowledge_review.py +65 -2
- package/graph-provider/fuli_graph/knowledge_review_models.py +5 -0
- package/graph-provider/fuli_graph/knowledge_search.py +0 -31
- package/graph-provider/fuli_graph/knowledge_versions.py +12 -1
- package/graph-provider/fuli_graph/models.py +9 -2
- package/graph-provider/fuli_graph/preference_versions.py +28 -7
- package/graph-provider/fuli_graph/project_agent_routes.py +60 -2
- package/graph-provider/fuli_graph/project_agent_task_models.py +63 -5
- package/graph-provider/fuli_graph/store.py +2 -0
- package/graph-provider/fuli_graph/store_project_agent_task_activity.py +12 -3
- package/graph-provider/fuli_graph/store_project_agent_task_blockers.py +110 -23
- package/graph-provider/fuli_graph/store_project_agent_task_pages.py +186 -0
- package/graph-provider/fuli_graph/store_project_agent_task_reads.py +6 -4
- package/graph-provider/fuli_graph/store_project_agent_task_recruitment.py +0 -7
- package/graph-provider/fuli_graph/store_project_agent_tasks.py +4 -3
- package/graph-provider/fuli_graph/store_task_contexts.py +14 -4
- package/graph-provider/fuli_graph/task_blocker.py +38 -6
- package/graph-provider/fuli_graph/tonborg_confirmation.py +48 -0
- package/graph-provider/pyproject.toml +1 -1
- package/native/win32/fuli-win-helper.exe +0 -0
- package/npm-shrinkwrap.json +16 -50
- package/package.json +2 -2
- package/packages/roundtable/src/peer/node-store.js +2 -6
- package/packages/roundtable/src/roundtable/chat-service.js +2 -1
- package/packages/roundtable/src/roundtable/chat.js +11 -1
- package/packages/roundtable/src/roundtable/delivery-state.js +43 -0
- package/packages/roundtable/src/roundtable/fla-address.js +6 -0
- package/packages/roundtable/src/roundtable/group-discussion-store.js +2 -2
- package/packages/roundtable/src/roundtable/group-discussion.js +15 -15
- package/packages/roundtable/src/roundtable/person-conversation.js +2 -1
- package/packages/roundtable/src/roundtable/service.js +40 -9
- package/packages/roundtable/src/roundtable/store.js +30 -7
- package/packages/roundtable/src/storage/database-format.js +21 -16
- package/registry/client-links.json +66 -4
- package/registry/occupation-plugins.json +593 -0
- package/skills/capturing-session-knowledge/SKILL.md +16 -3
- package/skills/capturing-session-knowledge/references/capture.md +22 -6
- package/skills/capturing-session-knowledge/references/coordination.md +16 -6
- package/skills/capturing-session-knowledge/references/retrieval.md +23 -6
- package/skills/flreview/SKILL.md +1 -1
- package/skills/operating-fuli/SKILL.md +40 -13
- package/skills/operating-fuli/references/workflow-recipes.md +9 -5
- package/src/acceptance-checks/port.js +37 -0
- package/src/adaptive-runtime/settings.js +1 -1
- package/src/adaptive-runtime/state.js +3 -3
- package/src/agent-apps/host-tools/catalog.js +7 -0
- package/src/agent-apps/store.js +2 -10
- package/src/agent-delegation/person-words.js +15 -1
- package/src/agent-kits/cli-presence.js +78 -0
- package/src/agent-kits/client-installer.js +9 -5
- package/src/agent-kits/installer-manifest.js +11 -6
- package/src/agent-kits/legacy-hook-cleanup.js +4 -3
- package/src/agent-kits/new-sessions.js +82 -0
- package/src/agent-kits/opencode-plugin-gate.js +3 -1
- package/src/agent-kits/pins.js +12 -12
- package/src/agent-kits/ports.js +5 -2
- package/src/agent-kits/session-launcher.js +11 -4
- package/src/agent-kits/vendor/agent-installer/README.md +1 -1
- package/src/agent-kits/vendor/agent-installer/index.d.ts +1 -1
- package/src/agent-kits/vendor/agent-session-launcher/README.md +1 -1
- package/src/agent-kits/vendor/agent-session-launcher/index.d.ts +30 -1
- package/src/agent-kits/vendor/agent-session-launcher/index.mjs +8 -0
- package/src/agent-roundtable/tools.js +3 -3
- package/src/agent-services/store.js +2 -4
- package/src/agent-tools/agent-workflow-definitions.js +7 -3
- package/src/agent-tools/checkpoint-definitions.js +5 -4
- package/src/agent-tools/fla-lesson-definitions.js +24 -0
- package/src/agent-tools/graph-definitions.js +8 -0
- package/src/agent-tools/graph-handlers.js +15 -4
- package/src/agent-tools/interface-catalog.js +34 -9
- package/src/agent-tools/model-backend-definitions.js +1 -1
- package/src/agent-tools/occupation-kit-definitions.js +20 -0
- package/src/agent-tools/project-agent-definitions.js +19 -40
- package/src/agent-tools/project-agent-tool-definitions.js +5 -1
- package/src/agent-tools/reply-echo-definitions.js +18 -0
- package/src/agent-tools/tonborg-question-definitions.js +26 -0
- package/src/agent-workflows/agent-tools.js +69 -3
- package/src/agent-workflows/port.js +32 -4
- package/src/agent-workflows/store.js +16 -9
- package/src/agents/call-source.js +11 -0
- package/src/agents/choice-answers.js +19 -0
- package/src/agents/claude-code/choice-answers.js +49 -0
- package/src/agents/claude-code/fla-menu-command.js +14 -0
- package/src/agents/claude-code/lifecycle-hook.js +16 -9
- package/src/agents/claude-code/turn-records.js +38 -0
- package/src/agents/claude-desktop/code-tab-token-source.js +26 -0
- package/src/agents/codex/fla-prompts.js +98 -0
- package/src/agents/codex/lifecycle-hook.js +8 -7
- package/src/agents/cursor/lifecycle-hook.js +3 -2
- package/src/agents/identity-receipt.js +9 -7
- package/src/agents/lifecycle-failure.js +15 -5
- package/src/agents/lifecycle-message.js +4 -5
- package/src/agents/opencode/lifecycle-hook.js +12 -7
- package/src/agents/opencode/plugin.js +7 -2
- package/src/agents/token-usage-resolver.js +1 -1
- package/src/agents/turn-receipt.js +5 -4
- package/src/cli/command-registry.js +1 -1
- package/src/cli/local-runtime-command.js +5 -2
- package/src/cli/local-runtime-options.js +7 -3
- package/src/cli/models-command.js +15 -11
- package/src/cli/update-command.js +2 -2
- package/src/cli/update-installation.js +23 -6
- package/src/cli/workspace-connection-command.js +2 -2
- package/src/client-links/registry.js +6 -112
- package/src/conversations/current-task-lookup.js +2 -3
- package/src/conversations/entry-skeleton.js +113 -55
- package/src/conversations/human-interaction.js +1 -1
- package/src/conversations/person-words.js +29 -0
- package/src/conversations/preference-session.js +17 -14
- package/src/conversations/stop-auto-close.js +28 -7
- package/src/conversations/task-context-view.js +113 -27
- package/src/conversations/turn-findings.js +105 -0
- package/src/data-format/format-notices.js +2 -1
- package/src/data-format/local-stores.js +22 -1
- package/src/debug-turns/model-capture.js +1 -1
- package/src/debug-turns/record.js +1 -1
- package/src/desktop/cli-path.js +253 -0
- package/src/desktop/desktop-app-package.js +31 -0
- package/src/desktop/{fuli-links.js → desktop-client-links.js} +20 -16
- package/src/desktop/desktop-controller.js +35 -3
- package/src/desktop/npm-adoption-store.js +7 -0
- package/src/desktop/npm-adoption.js +119 -0
- package/src/desktop/npm-copies.js +73 -0
- package/src/desktop/npm-copy-control.js +56 -0
- package/src/desktop/npm-copy-data.js +51 -0
- package/src/desktop/npm-copy-uninstall.js +67 -0
- package/src/desktop/windows-controller.js +13 -1
- package/src/desktop/windows-installation-switch.js +1 -1
- package/src/desktop/windows-node-launcher.js +11 -0
- package/src/desktop/windows-uninstall.js +3 -0
- package/src/dreaming/astro-calls.js +47 -0
- package/src/dreaming/projects.js +39 -0
- package/src/dreaming/run.js +38 -36
- package/src/dreaming/scheduler.js +2 -2
- package/src/employees/builtin/jefa/board-store.js +2 -10
- package/src/employees/builtin/jefa/tools.js +1 -1
- package/src/employees/management-policy.js +1 -1
- package/src/employees/manifest.js +1 -1
- package/src/employees/service.js +11 -1
- package/src/external-knowledge/document-mapping.js +1 -1
- package/src/external-knowledge/index.js +0 -4
- package/src/fla-growth/lessons.js +40 -0
- package/src/fla-growth/task-verification.js +25 -0
- package/src/fla-growth/tool-suggestions.js +107 -0
- package/src/fla-growth/verification.js +30 -0
- package/src/graphiti/agent-access-policy.js +1 -1
- package/src/graphiti/agent-conversations.js +2 -2
- package/src/graphiti/agent-growth.js +32 -4
- package/src/graphiti/capture-change-confirmation.js +4 -1
- package/src/graphiti/capture-notice.js +2 -1
- package/src/graphiti/capture-onboarding.js +1 -1
- package/src/graphiti/checkpoint-provider-outage.js +23 -0
- package/src/graphiti/collaboration-preference-projection.js +1 -1
- package/src/graphiti/conversation-triage.js +1 -1
- package/src/graphiti/execution-mode.js +1 -1
- package/src/graphiti/federated-application.js +16 -8
- package/src/graphiti/fla-mention.js +105 -96
- package/src/graphiti/fla-menu-reader.js +25 -0
- package/src/graphiti/fla-menu.js +108 -0
- package/src/graphiti/fla-picker.js +21 -16
- package/src/graphiti/fuli-markup.js +60 -0
- package/src/graphiti/group-member-context.js +1 -1
- package/src/graphiti/human-review-channel.js +16 -2
- package/src/graphiti/knowledge-result-budget.js +2 -2
- package/src/graphiti/knowledge-review.js +2 -1
- package/src/graphiti/local-runtimes.js +11 -0
- package/src/graphiti/preference-task-scope.js +1 -1
- package/src/graphiti/project-agent-control-plane.js +6 -0
- package/src/graphiti/project-agent-mapping.js +23 -10
- package/src/graphiti/project-agent-mention.js +10 -28
- package/src/graphiti/project-agent-task-entry.js +1 -1
- package/src/graphiti/project-agent-task-sheets.js +1 -1
- package/src/graphiti/project-agent-workflows.js +20 -5
- package/src/graphiti/project-knowledge-presence.js +27 -0
- package/src/graphiti/project-path-context.js +2 -2
- package/src/graphiti/provider-client.js +24 -0
- package/src/graphiti/provider-outage.js +20 -0
- package/src/graphiti/recall-policy.js +1 -1
- package/src/graphiti/reply-echo-agent-tool.js +36 -0
- package/src/graphiti/reply-echo-group.js +85 -0
- package/src/graphiti/reply-echo.js +87 -0
- package/src/graphiti/session-knowledge-migration.js +2 -1
- package/src/graphiti/session-prompt-sentences.js +52 -0
- package/src/graphiti/session-records.js +3 -2
- package/src/graphiti/source-marker.js +14 -0
- package/src/graphiti/task-blocker-progress.js +24 -1
- package/src/graphiti/task-context-entry.js +20 -1
- package/src/graphiti/task-entry-group-discussion.js +31 -31
- package/src/graphiti/task-entry-named-fla.js +40 -70
- package/src/graphiti/task-entry-preferences.js +2 -2
- package/src/graphiti/task-entry-project-scope.js +4 -4
- package/src/graphiti/task-entry-provider-outage.js +25 -0
- package/src/graphiti/task-entry-read-only.js +1 -1
- package/src/graphiti/task-entry-standing-owner.js +3 -3
- package/src/graphiti/task-experiences.js +21 -3
- package/src/graphiti/task-knowledge-checkpoint.js +5 -2
- package/src/graphiti/task-knowledge-recall.js +140 -25
- package/src/graphiti/task-prompt-statement.js +49 -0
- package/src/graphiti/task-remember.js +23 -9
- package/src/graphiti/task-scene.js +1 -1
- package/src/graphiti/task-stated-rules.js +75 -0
- package/src/graphiti/temporary-project-scope.js +1 -1
- package/src/graphiti/topic-proposals.js +6 -5
- package/src/graphiti/user-taste-skill.js +1 -1
- package/src/graphiti/workflow-candidate-mapping.js +1 -1
- package/src/http/agent-delegation-gate.js +9 -0
- package/src/http/api-router.js +16 -0
- package/src/http/cli-path-api-router.js +35 -0
- package/src/http/desktop-client-api-router.js +18 -0
- package/src/http/dreaming-api-router.js +10 -7
- package/src/http/graph-api-router.js +17 -0
- package/src/http/notices-api-router.js +37 -0
- package/src/http/npm-copies-api-router.js +36 -0
- package/src/http/occupation-plugins-api-router.js +43 -0
- package/src/http/reply-echo-api-router.js +53 -0
- package/src/http/response.js +1 -1
- package/src/http/retrospectives-api-router.js +18 -0
- package/src/http/session-launch-api-router.js +34 -1
- package/src/http/static-handler.js +1 -1
- package/src/http/task-blocker-api-router.js +17 -5
- package/src/http/task-verification-api-router.js +16 -0
- package/src/http/whats-new-api-router.js +34 -0
- package/src/judgment/acceptance.js +10 -3
- package/src/judgment/agent-view.js +21 -4
- package/src/judgment/conversation.js +86 -2
- package/src/judgment/delegated-reviews.js +23 -0
- package/src/judgment/engine.js +49 -16
- package/src/judgment/explain-method.js +11 -0
- package/src/judgment/http.js +17 -2
- package/src/judgment/question-options.js +97 -0
- package/src/judgment/question-tools.js +235 -0
- package/src/judgment/review-delegation.js +89 -0
- package/src/judgment/review-prompt.js +44 -3
- package/src/judgment/review-verdicts.js +13 -4
- package/src/judgment/runtime.js +3 -1
- package/src/judgment/service.js +25 -5
- package/src/judgment/store.js +44 -8
- package/src/judgment/verdict-schema.js +6 -0
- package/src/judgment/visible-items.js +15 -4
- package/src/knowledge-duplicates/scheduler.js +28 -0
- package/src/knowledge-hard-delete/forget-references.js +5 -2
- package/src/knowledge-impressions/store.js +17 -5
- package/src/local-runtime/console-autostart-record.js +95 -0
- package/src/local-runtime/console-autostart.js +143 -0
- package/src/local-runtime/console-health.js +26 -0
- package/src/local-runtime/installed-console-launcher.js +73 -0
- package/src/local-runtime/lifecycle.js +85 -21
- package/src/local-runtime/processes.js +15 -2
- package/src/local-runtime/unresponsive-console.js +77 -0
- package/src/mcp/create-mcp-server.js +11 -4
- package/src/mcp/fla-menu-prompts.js +60 -0
- package/src/mcp/runtime.js +11 -3
- package/src/mcp/session-id.js +3 -2
- package/src/mcp/tool-annotations.js +5 -2
- package/src/mcp/tool-profiles.js +1 -1
- package/src/mcp/tool-result.js +3 -1
- package/src/model-backends/adapters/anthropic-messages.js +2 -2
- package/src/model-backends/adapters/ollama-local-build.js +1 -1
- package/src/model-backends/adapters/openai-chat.js +2 -2
- package/src/model-backends/budget.js +10 -1
- package/src/model-backends/config.js +2 -2
- package/src/model-backends/endpoint-models.js +80 -0
- package/src/model-backends/http.js +3 -1
- package/src/model-backends/keys-http.js +5 -4
- package/src/model-backends/models-http.js +54 -0
- package/src/model-backends/port.js +32 -5
- package/src/model-backends/prompt.js +11 -1
- package/src/model-backends/role-settings.js +27 -2
- package/src/model-backends/role-trust.js +3 -15
- package/src/model-backends/role-work-store.js +2 -12
- package/src/model-backends/runtime.js +11 -2
- package/src/model-backends/secret-store.js +61 -12
- package/src/model-backends/suggestion-log.js +6 -15
- package/src/model-backends/usage-ledger.js +28 -18
- package/src/notices/recent-notices.js +175 -0
- package/src/notifications/needs-person.js +1 -1
- package/src/occupation-plugins/agent-tools.js +59 -0
- package/src/occupation-plugins/choice-store.js +65 -0
- package/src/occupation-plugins/manifest.js +165 -0
- package/src/occupation-plugins/port.js +104 -0
- package/src/occupation-plugins/recruitment.js +38 -0
- package/src/presence/windows-helper-pin.js +1 -1
- package/src/registry/updatable-manifest.js +127 -0
- package/src/retrieval-terms/fill.js +1 -1
- package/src/retrospectives/run.js +113 -0
- package/src/retrospectives/runs.js +18 -0
- package/src/retrospectives/scheduler.js +58 -0
- package/src/server/application-lifecycle.js +1 -1
- package/src/server/signal-shutdown.js +31 -0
- package/src/server.js +26 -18
- package/src/setup/first-run-capture.js +6 -2
- package/src/setup/graph-runtime.js +46 -38
- package/src/setup/neo4j-memory-profile.js +1 -1
- package/src/setup/paths.js +2 -0
- package/src/standing-assignments/store.js +2 -4
- package/src/system/client-connection-status.js +3 -1
- package/src/system/doctor.js +19 -2
- package/src/system/graph-migration.js +1 -1
- package/src/system/platform-path.js +21 -9
- package/src/system/runtime-settings.js +19 -7
- package/src/system/settings-guide.js +7 -0
- package/src/system/system-service.js +41 -2
- package/src/task-blockers/console-port.js +12 -6
- package/src/text/markdown-link.js +12 -0
- package/src/text/single-line.js +6 -0
- package/src/text/utf8-preview.js +7 -6
- package/src/whats-new/lines/0.25.0.json +17 -0
- package/src/whats-new/lines/0.25.1.json +13 -0
- package/src/whats-new/release-lines.js +66 -0
- package/src/whats-new/whats-new.js +70 -0
- package/src/work-surfaces/store.js +2 -4
- package/dist/web/assets/AgentAppsPage-BYXraPFM.js +0 -1
- package/dist/web/assets/AgentManagement-Bza8k5ra.js +0 -7
- package/dist/web/assets/AgentManagement-CeF4mjtg.css +0 -1
- package/dist/web/assets/AgentMentionCopy-C72hY2QL.js +0 -1
- package/dist/web/assets/AgentMentionCopy-CNKrzZvE.css +0 -1
- package/dist/web/assets/AgentProfilePage-kp7-gRRx.js +0 -5
- package/dist/web/assets/AgentServicesPage-ARIWfmuh.css +0 -1
- package/dist/web/assets/AgentServicesPage-IsGaz7FY.js +0 -1
- package/dist/web/assets/ConnectionsPage-BpvecWXD.js +0 -1
- package/dist/web/assets/ConnectionsPage-Cwt-knQQ.css +0 -1
- package/dist/web/assets/DataFormatPage-dsy8d3oA.js +0 -2
- package/dist/web/assets/DataFormatPage-sSnjjUY4.css +0 -1
- package/dist/web/assets/DebugTurnPage-cIJV7xDW.js +0 -1
- package/dist/web/assets/DebugTurnsPage-B6Te_Alp.js +0 -1
- package/dist/web/assets/EmployeeScopeDialog-CXNZ62Ab.js +0 -1
- package/dist/web/assets/EmployeeScopeDialog-Do5PHQh7.css +0 -1
- package/dist/web/assets/EmployeeWorkbench-DALIO-cM.js +0 -4
- package/dist/web/assets/FieldRoot-DRnAC66O.js +0 -1
- package/dist/web/assets/FlaGrowth-7wlv31Tl.js +0 -1
- package/dist/web/assets/FlaGrowth-BzXY8pcx.css +0 -1
- package/dist/web/assets/FlasPage-BCs1Ffpf.js +0 -1
- package/dist/web/assets/FlasPage-SLQDMv47.css +0 -1
- package/dist/web/assets/InboxPage-Bkx6nIhb.css +0 -1
- package/dist/web/assets/InboxPage-CgJnymOY.js +0 -4
- package/dist/web/assets/KnowledgeItemEditPage-rwpM8cF1.js +0 -1
- package/dist/web/assets/KnowledgeOrganizerPage-7IdnjM3L.js +0 -1
- package/dist/web/assets/KnowledgeOrganizerPage-CNm2jvJZ.css +0 -1
- package/dist/web/assets/KnowledgePage-gxqOIZxf.js +0 -1
- package/dist/web/assets/KnowledgePeek-CQyCxHFj.js +0 -1
- package/dist/web/assets/KnowledgeWorkspace-A7zVb26p.js +0 -4
- package/dist/web/assets/MenuSeparator-quqkZZXY.js +0 -1
- package/dist/web/assets/ModelSettingsPage-B8jC_uew.js +0 -3
- package/dist/web/assets/ModelSettingsPage-DhCn_SF6.css +0 -1
- package/dist/web/assets/OccupationsPage-BVREfX_R.js +0 -1
- package/dist/web/assets/OccupationsPage-Di7B-bPl.css +0 -1
- package/dist/web/assets/OverviewPage-CHyEbdP4.js +0 -1
- package/dist/web/assets/OverviewPage-xdF5g6bZ.css +0 -1
- package/dist/web/assets/PageOutline-BKBQ1T4d.js +0 -1
- package/dist/web/assets/PageOutline-BqVgnlos.css +0 -1
- package/dist/web/assets/PersonalProfilePage-Cxp2gn6J.js +0 -1
- package/dist/web/assets/PersonalProjectsPage-821imOxR.js +0 -1
- package/dist/web/assets/PopperContent-DNqbLkNa.js +0 -1
- package/dist/web/assets/ProfileAuthorChip-7HSHODdR.js +0 -1
- package/dist/web/assets/ProfileAuthorChip-C_0hzDoP.css +0 -1
- package/dist/web/assets/ProjectAgentsPage-DBsXS6Xa.js +0 -5
- package/dist/web/assets/RovingFocusGroup-BB0d3k17.js +0 -1
- package/dist/web/assets/SessionRecordsPage-BMAAtIAt.js +0 -1
- package/dist/web/assets/SettingsPage-BYNEk6BM.js +0 -1
- package/dist/web/assets/SettingsPage-Uj1IpZpS.css +0 -1
- package/dist/web/assets/SharedJefaReportPage-g35o6jUb.js +0 -1
- package/dist/web/assets/TagVocabularyPage-COubNJXz.js +0 -1
- package/dist/web/assets/TaskBlockerClear-B5MpHdlh.js +0 -1
- package/dist/web/assets/TaskBlockerClear-CjSNfpyP.css +0 -1
- package/dist/web/assets/TaskBlockerNotice-CIN0HRiE.js +0 -1
- package/dist/web/assets/TeamCard-CsmWDw2Q.css +0 -1
- package/dist/web/assets/TeamCard-D-kLa5rT.js +0 -1
- package/dist/web/assets/TeamIsland-BIo-l49b.js +0 -9
- package/dist/web/assets/UiIconButton-CCo4NPYv.js +0 -1
- package/dist/web/assets/UiNotice-DGPFbVo7.js +0 -1
- package/dist/web/assets/VirtualDirectoryList-1vLi41SD.css +0 -1
- package/dist/web/assets/VirtualDirectoryList-BkV3gj5E.js +0 -1
- package/dist/web/assets/VisuallyHiddenInput-Bdcr9CwC.js +0 -1
- package/dist/web/assets/WidgetSlot-DFojQaSo.js +0 -2
- package/dist/web/assets/_plugin-vue_export-helper-B9AJmZbc.js +0 -9
- package/dist/web/assets/agent-apps-CKGFfySp.js +0 -1
- package/dist/web/assets/agent-avatar-V6M_DF3r.js +0 -1
- package/dist/web/assets/agent-occupation-PZ2YAIM8.js +0 -1
- package/dist/web/assets/agent-profile-Clb-gYu-.css +0 -1
- package/dist/web/assets/agent-work-summary-bzTwEUaA.js +0 -1
- package/dist/web/assets/checkbox-DPyLHneg.js +0 -1
- package/dist/web/assets/client-connection-model-vO2cap6a.js +0 -1
- package/dist/web/assets/client-links-store-DMqe0MbK.js +0 -1
- package/dist/web/assets/combobox-zUq9bkyM.js +0 -1
- package/dist/web/assets/console-DaoAO-EO.js +0 -1
- package/dist/web/assets/context-menu-CcGDnF_k.js +0 -1
- package/dist/web/assets/dropdown-menu-CXmMnCBh.js +0 -1
- package/dist/web/assets/fla-mention-D_Qb7VeY.js +0 -1
- package/dist/web/assets/index-B4AMPpl_.css +0 -1
- package/dist/web/assets/index-BYIOAXnn.js +0 -2
- package/dist/web/assets/label-table-BnWNYfyk.js +0 -1
- package/dist/web/assets/model-CDevUwUw.js +0 -1
- package/dist/web/assets/model-P5fSGsiJ.js +0 -1
- package/dist/web/assets/motion-QY5FKKko.js +0 -1
- package/dist/web/assets/occupation-catalog-uDDhZwKK.js +0 -1
- package/dist/web/assets/person-profile-DXXfpvta.js +0 -1
- package/dist/web/assets/popup-host-Bmvrf-2t.js +0 -1
- package/dist/web/assets/profile-model-6gTmScHC.js +0 -1
- package/dist/web/assets/quick-edit-api-WLh_88w6.js +0 -1
- package/dist/web/assets/radio-group-B4r78nDh.js +0 -1
- package/dist/web/assets/roundtable-live-DFeDF56c.js +0 -1
- package/dist/web/assets/save-file-rPduHi2f.js +0 -1
- package/dist/web/assets/select-_JO3Mxdp.js +0 -1
- package/dist/web/assets/tabs-BlcZiEMY.js +0 -1
- package/dist/web/assets/ui-BopQuQfc.js +0 -1
- package/dist/web/assets/ui-CoUaCzmO.css +0 -1
- package/dist/web/assets/useAgentWorkSummary-BQbhaHov.js +0 -1
- package/dist/web/assets/useForwardScopeId-CUQFIAIj.js +0 -1
- package/dist/web/assets/useKnowledgeItemPeek-PmQJixw5.js +0 -1
- package/dist/web/assets/useTeamPeople-CDISP0PM.js +0 -1
- package/dist/web/assets/utils-B4pBTmyh.js +0 -1
- package/src/workspace-protocol/array-validation.js +0 -19
- package/src/workspace-protocol/base64url-contract.js +0 -27
- package/src/workspace-protocol/constants.js +0 -37
- package/src/workspace-protocol/error-contract.js +0 -24
- package/src/workspace-protocol/fact-contract.js +0 -110
- package/src/workspace-protocol/governance-contract.js +0 -47
- package/src/workspace-protocol/index.js +0 -53
- package/src/workspace-protocol/json-budget.js +0 -19
- package/src/workspace-protocol/manifest-contract.js +0 -186
- package/src/workspace-protocol/publication-contract.js +0 -212
- package/src/workspace-protocol/query-contract.js +0 -149
- package/src/workspace-protocol/scalar-contract.js +0 -192
- package/src/workspace-protocol/sync-contract.js +0 -184
- package/src/workspace-protocol/text-contract.js +0 -39
- package/src/workspace-protocol/validation.js +0 -66
- /package/dist/web/assets/{UiNotice-ZCBdABdM.css → UiIconButton-ZCBdABdM.css} +0 -0
- /package/dist/web/assets/{settings-restart-BU_GGtbG.css → settings-autosave-BU_GGtbG.css} +0 -0
package/README.md
CHANGED
|
@@ -1,1033 +1,223 @@
|
|
|
1
1
|
<p align="center">
|
|
2
2
|
<picture>
|
|
3
|
-
<source media="(prefers-color-scheme: dark)" srcset="
|
|
4
|
-
<img src="
|
|
3
|
+
<source media="(prefers-color-scheme: dark)" srcset="brand/svg/lockup/fuli-lockup-bilingual-horizontal-dark.svg" />
|
|
4
|
+
<img src="brand/svg/lockup/fuli-lockup-bilingual-horizontal.svg" alt="FULI 复利 logo" width="200" />
|
|
5
5
|
</picture>
|
|
6
6
|
</p>
|
|
7
7
|
|
|
8
|
-
<
|
|
8
|
+
<p align="center">
|
|
9
|
+
<picture>
|
|
10
|
+
<source media="(prefers-color-scheme: dark)" srcset="docs/readme/hero-en-dark.webp" />
|
|
11
|
+
<img src="docs/readme/hero-en.webp" alt="FULI 复利: your AI tools share one memory of you and your projects" width="800" />
|
|
12
|
+
</picture>
|
|
13
|
+
</p>
|
|
14
|
+
|
|
15
|
+
<p align="center">FULI is a plugin that lets your tasks move across all your Agent clients: Claude Code, Codex, OpenCode… They share one memory, one taste and one FLA team, so you use whichever fits and switch without explaining everything again. It doesn't replace your clients or change what you set up: your models, subscriptions, sessions and settings stay as they are. FULI plugs in as its own entries only, and removing it from a client in the console, or uninstalling FULI, takes those entries out again.</p>
|
|
16
|
+
|
|
9
17
|
|
|
10
18
|
<p align="center">
|
|
11
|
-
English · <a href="README.zh-CN.md">简体中文</a>
|
|
19
|
+
English · <a href="README.zh-CN.md">简体中文</a> · <a href="https://有点.意思.com/fuli/">Website</a> · <a href="https://github.com/JevonsCode/fuli/releases/latest">Download</a>
|
|
12
20
|
</p>
|
|
13
21
|
|
|
14
22
|
<p align="center">
|
|
23
|
+
<a href="https://github.com/JevonsCode/fuli/releases/latest"><img src="https://img.shields.io/github/v/release/JevonsCode/fuli?style=flat&label=desktop" alt="latest release" /></a>
|
|
15
24
|
<a href="https://www.npmjs.com/package/fuli-context"><img src="https://img.shields.io/npm/v/fuli-context?style=flat&logo=npm&label=fuli-context" alt="npm version" /></a>
|
|
16
|
-
<a href="https://
|
|
25
|
+
<a href="https://github.com/JevonsCode/fuli/releases"><img src="https://img.shields.io/github/downloads/JevonsCode/fuli/total?style=flat&label=desktop%20downloads" alt="desktop downloads" /></a>
|
|
26
|
+
<a href="https://www.npmjs.com/package/fuli-context"><img src="https://img.shields.io/npm/dm/fuli-context?style=flat&label=npm%20downloads" alt="npm downloads" /></a>
|
|
17
27
|
<a href="LICENSE"><img src="https://img.shields.io/npm/l/fuli-context?style=flat" alt="license" /></a>
|
|
18
28
|
</p>
|
|
19
29
|
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
|
|
23
|
-
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
|
|
30
|
-
|
|
31
|
-
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
|
|
35
|
-
|
|
36
|
-
|
|
37
|
-
|
|
38
|
-
|
|
39
|
-
|
|
40
|
-
|
|
41
|
-
|
|
42
|
-
|
|
43
|
-
|
|
44
|
-
|
|
45
|
-
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
|
|
49
|
-
|
|
50
|
-
|
|
51
|
-
|
|
52
|
-
|
|
53
|
-
|
|
54
|
-
|
|
55
|
-
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
|
|
63
|
-
|
|
64
|
-
|
|
65
|
-
|
|
66
|
-
|
|
67
|
-
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
77
|
-
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
82
|
-
|
|
83
|
-
no administrator rights; Node.js, Java and the graph database are built in. On first launch FULI
|
|
84
|
-
finds the AI clients on your computer and connects them in one click, and offers start at login.
|
|
85
|
-
If FULI is already installed, the setup updates it and keeps your data. The installer is not
|
|
86
|
-
code-signed yet, so Windows may show "Windows protected your PC": choose "More info → Run anyway".
|
|
87
|
-
- **Download the app (macOS, Apple silicon):** get `fuli-macos-arm64-<version>.dmg` from the same
|
|
88
|
-
release, open it and drag FULI.app to Applications. macOS 13.5 or later; Intel Macs use the terminal
|
|
89
|
-
install. The app is not signed yet: the first time it is blocked, open System Settings › Privacy &
|
|
90
|
-
Security and choose "Open Anyway". See [docs/macos-app.md](docs/macos-app.md).
|
|
91
|
-
- **Install from the terminal:** for developers, and for macOS / Linux.
|
|
92
|
-
|
|
93
|
-
With Node.js 24.12+ and Docker Compose v2 available (default container mode; see
|
|
94
|
-
[Installation](#installation) for macOS/Linux native mode), run:
|
|
95
|
-
|
|
96
|
-
```bash
|
|
97
|
-
npm install --global fuli-context
|
|
98
|
-
fuli setup
|
|
99
|
-
fuli open
|
|
100
|
-
```
|
|
101
|
-
|
|
102
|
-
The setup wizard shows its plan before changing your environment.
|
|
103
|
-
|
|
104
|
-
### 2. Connect a client
|
|
105
|
-
|
|
106
|
-
- **Detected Codex, Claude Code, Cursor, or Claude Desktop:** `fuli setup` writes the MCP entry and
|
|
107
|
-
supported Hooks/Skills. Quit and reopen apps that read MCP configuration only at launch, then open
|
|
108
|
-
a new task. Codex Hook trust requires manual `/hooks` review in Codex.
|
|
109
|
-
- **Other MCP clients:** run `fuli connect` (or open **Settings → Client connection** in the
|
|
110
|
-
console), merge the `fuli` entry into the client's MCP configuration while keeping other servers,
|
|
111
|
-
and reload or trust the server in that client. `fuli connect` prints JSON only.
|
|
112
|
-
|
|
113
|
-
### 3. First actual tool call and verification
|
|
114
|
-
|
|
115
|
-
1. Run `fuli status` and confirm the management UI and personal graph are ready (default
|
|
116
|
-
`http://127.0.0.1:2727`).
|
|
117
|
-
2. Confirm Fuli tools appear in the client and **make one actual tool call** inside a project
|
|
118
|
-
directory (for example, ask the Agent to load project collaboration preferences or coordinate a
|
|
119
|
-
small task). Verify that Fuli resolves the right project (or registers/staffs a lead Agent) and
|
|
120
|
-
asks for missing context rather than inventing it.
|
|
121
|
-
3. Optional cross-tool handoff: use the Agent's continuation instruction in another client
|
|
122
|
-
connected to the same data and verify that it recovers the same Agent, project, and prior work.
|
|
123
|
-
See [Agent conversations and collaboration](docs/agent-conversations-and-collaboration.md).
|
|
124
|
-
|
|
125
|
-
For a one-off task, this source version also supports an isolated **temporary project** when no
|
|
126
|
-
registered project matches; reading preferences alone creates nothing, and an ambiguous project
|
|
127
|
-
still needs a choice. Fuli retains its task history and follows the existing Agent recruitment and
|
|
128
|
-
executor permissions. The Agent directory explains the first-task entry, and task details show a
|
|
129
|
-
completion table only after real worker execution is reported — an assigned Agent or an ordinary
|
|
130
|
-
local conversation alone is not worker evidence.
|
|
131
|
-
|
|
132
|
-
### 4. Troubleshooting
|
|
133
|
-
|
|
134
|
-
If setup, client detection, or tool calls fail, see
|
|
135
|
-
[Common failure recovery](docs/client-connections.md#common-failure-recovery). Use
|
|
136
|
-
`fuli setup --yes --connected-only` only to refresh clients that still have an existing Fuli MCP
|
|
137
|
-
entry (**Connected** / `connected` or **Configuration changed** / `changed` in Settings); clients
|
|
138
|
-
showing **Skills only** (`skills_left`), **Removed** (`removed`), **Not connected**
|
|
139
|
-
(`not_connected`), or newly installed clients without a Fuli MCP entry require a full `fuli setup`.
|
|
140
|
-
|
|
141
|
-
[How it works](#philosophy) · [Installation](#installation) ·
|
|
142
|
-
[Agent roles and teams](docs/employee-agents.md) ·
|
|
143
|
-
[Cross-client acceptance](acceptance/cross-client-handoff.md) · [Contribute](CONTRIBUTING.md) ·
|
|
144
|
-
[Report an issue](https://github.com/JevonsCode/fuli/issues/new/choose)
|
|
145
|
-
|
|
146
|
-
## Agent Roundtable
|
|
147
|
-
|
|
148
|
-
Agents talk to each other directly, the way a lead asks a teammate or a sub-agent. A Codex
|
|
149
|
-
Agent can ask the Agent behind a specific Claude Code conversation, or a FULI Agent that has
|
|
150
|
-
never run yet. Nobody creates rooms or invites participants: an Agent calls `message_agent`,
|
|
151
|
-
Fuli wakes the recipient read-only in its own client and returns its answer: Claude Code
|
|
152
|
-
forks the conversation; Codex records the exchange in it (or uses a new session while it is
|
|
153
|
-
open in the app). A message that cannot be delivered waits in the recipient's inbox for its
|
|
154
|
-
next task.
|
|
155
|
-
|
|
156
|
-
Agents can also keep direct and group chats with `agent_chat`. Messages are pushed: an Agent
|
|
157
|
-
that calls `wait` gets new messages at once, and a direct message or an `@` mention wakes an
|
|
158
|
-
Agent that is not waiting. Chats only carry communication; they never change knowledge weights.
|
|
159
|
-
|
|
160
|
-
Every exchange is kept locally; the console's Roundtable page shows who asked whom, through
|
|
161
|
-
which client, and exactly what was said. Tools and limits: [Agent Roundtable](docs/agent-roundtable.md).
|
|
162
|
-
|
|
163
|
-
## npm packages
|
|
164
|
-
|
|
165
|
-
| Package | Status | Intended use |
|
|
166
|
-
| --- | --- | --- |
|
|
167
|
-
| [`fuli-context`](https://www.npmjs.com/package/fuli-context) | Available | Personal edition; personal knowledge, the graph database, and the management UI all run locally |
|
|
168
|
-
| Fuli Server npm package | In development | A separate server for team-shared deployments; not yet published |
|
|
169
|
-
|
|
170
|
-
`fuli-context` is currently the only available npm package. The shared layer never contains
|
|
171
|
-
personal taste, personality, or judgment preferences.
|
|
172
|
-
|
|
173
|
-
## Philosophy
|
|
174
|
-
|
|
175
|
-
### 1. Measure reuse value, not knowledge volume
|
|
176
|
-
|
|
177
|
-
Fuli is built around one question:
|
|
178
|
-
|
|
179
|
-
> Without Fuli, would the agent have to relearn information that is confirmed, still valid, and
|
|
180
|
-
> relevant to the current task?
|
|
181
|
-
|
|
182
|
-
Node count, transcript count, and graph size are not success metrics. The product succeeds when a
|
|
183
|
-
past collaboration asset is reused within the correct project, time, provenance, and authority
|
|
184
|
-
boundaries—and reduces repeated explanation, correction, or rework.
|
|
185
|
-
|
|
186
|
-
### 2. Review every task; do not retain every task
|
|
187
|
-
|
|
188
|
-
A task may produce:
|
|
189
|
-
|
|
190
|
-
- reusable knowledge such as API conventions, runbooks, release procedures, or constraints;
|
|
191
|
-
- experience explaining why an approach worked or failed;
|
|
192
|
-
- a decision trace with options, a human choice, rationale, and later validation;
|
|
193
|
-
- judgment aids such as preferences, principles, taste, or personality tendencies.
|
|
194
|
-
|
|
195
|
-
It may also produce only disposable output, a guess, or a one-off choice. In that case,
|
|
196
|
-
`retain_nothing` is the correct result. Every task should perform a knowledge-value checkpoint, but
|
|
197
|
-
not every task must become stored knowledge.
|
|
198
|
-
|
|
199
|
-
### 3. Human authority outranks agent repetition
|
|
200
|
-
|
|
201
|
-
An agent may discover candidates, surface conflicts, recommend promotion, or resolve a deferred
|
|
202
|
-
conflict when explicitly authorized. It may not turn semantic similarity, repetition, or its own
|
|
203
|
-
judgment into human confirmation. Shared-knowledge promotion and sensitive project mutations use a
|
|
204
|
-
preview followed by an explicit, one-time, atomic action.
|
|
205
|
-
|
|
206
|
-
### 4. Assemble context, then retrieve evidence on demand
|
|
207
|
-
|
|
208
|
-
Fuli does not inject the complete history at session start. The entry step always loads effective
|
|
209
|
-
personal-global preferences, preferences from one exactly matched project, and explicitly
|
|
210
|
-
propagating preferences from authorized parent projects. When the current task signals that a
|
|
211
|
-
stable prior fact, method, URL, decision, release, deployment, or authentication runbook may matter,
|
|
212
|
-
the same step also performs a small bounded recall from that project and its authorized knowledge
|
|
213
|
-
sources. The agent uses focused on-demand search when more evidence is needed.
|
|
214
|
-
Retrieval alone is not usage; a usage event is recorded only when an item materially affects an
|
|
215
|
-
answer, implementation, or decision.
|
|
216
|
-
|
|
217
|
-
### 5. Preserve evolution instead of overwriting history
|
|
218
|
-
|
|
219
|
-
An old approach may have been correct under earlier conditions and later replaced. Fuli preserves
|
|
220
|
-
provenance, confirmation authority, time, rationale, revisions, supersession, and negative evidence
|
|
221
|
-
so an agent can explain both what applies now and why the past was different. Negative feedback may
|
|
222
|
-
lower rank or trigger review, but it never silently erases history.
|
|
223
|
-
|
|
224
|
-
### 6. Put shared knowledge in the parent; keep differences in each child
|
|
225
|
-
|
|
226
|
-
For example, hotel and flight projects can keep their own PRDs, configuration, and domain rules,
|
|
227
|
-
while a parent activity-platform project owns shared local-run, mock, test, and deployment runbooks:
|
|
228
|
-
|
|
229
|
-
```text
|
|
230
|
-
Activity platform (parent: shared runbooks)
|
|
231
|
-
├── Hotel project (child: hotel PRD / configuration / overrides)
|
|
232
|
-
└── Flight project (child: flight PRD / configuration / overrides)
|
|
233
|
-
```
|
|
234
|
-
|
|
235
|
-
When an agent works in the hotel project, it searches hotel-local knowledge first, then follows
|
|
236
|
-
explicit outgoing `PART_OF` or `USES_KNOWLEDGE_FROM` relationships to authorized sources, up to two
|
|
237
|
-
hops. A child item with the same stable key overrides an inherited item. Project-scoped personal
|
|
238
|
-
preferences remain exact by default; only preferences explicitly marked `descendants` or
|
|
239
|
-
`selected_projects` may propagate over the same authorized relationships, with origin project,
|
|
240
|
-
path, and distance preserved. Equally near parent conflicts require human judgment and are never
|
|
241
|
-
resolved by weight. Generic `RELATED_TO` relationships do not expand scope automatically; search
|
|
242
|
-
returns only a one-time read-only expansion suggestion that the Agent must ask the user to approve.
|
|
243
|
-
|
|
244
|
-
Similar items from multiple children create only a common-knowledge candidate. A shared preference
|
|
245
|
-
across projects without one common parent likewise creates only a personal-global candidate while
|
|
246
|
-
preserving each source text, qualifier, and provenance. Current clustering is a lexical heuristic,
|
|
247
|
-
not proof of semantic equivalence. A human must choose the content, scope, and rationale before it
|
|
248
|
-
can apply.
|
|
249
|
-
|
|
250
|
-
### 7. Local first; no personal model in the shared layer
|
|
251
|
-
|
|
252
|
-
The personal graph stays local by default. Supported client adapters can journal visible
|
|
253
|
-
conversations for private agent continuity, separately from structured confirmed knowledge.
|
|
254
|
-
Conversation recall is bounded, with a default seven-day inactivity window. Credentials, temporary
|
|
255
|
-
logs, and raw command output must not enter the knowledge graph. The team-shared layer contains only
|
|
256
|
-
confirmed project or domain knowledge with context and provenance—not personal taste, personality,
|
|
257
|
-
or judgment preferences.
|
|
258
|
-
|
|
259
|
-
## Practical use: classification, scope, project resolution, and retrieval
|
|
260
|
-
|
|
261
|
-
### Decide what to store before deciding where to store it
|
|
262
|
-
|
|
263
|
-
Content type and effective scope are independent dimensions. The first three rows are personal
|
|
264
|
-
collaboration preferences and carry a `profileAspect`. Project facts have no `profileAspect` and do
|
|
265
|
-
not appear on the Personal preferences page.
|
|
266
|
-
|
|
267
|
-
| Type | What belongs here | Example |
|
|
268
|
-
| --- | --- | --- |
|
|
269
|
-
| Taste (`taste`) | Outcomes, styles, and quality directions the user explicitly likes or rejects across UI, writing, product, architecture, or engineering work | “Use fewer large gradients and avoid cards nested inside cards.” |
|
|
270
|
-
| Personality (`personality`) | A stable working or collaboration trait explicitly described by the user | “I communicate directly and want the Agent to move work forward proactively.” |
|
|
271
|
-
| Judgment preference (`judgment_preference`) | Decision conditions, priorities, risk posture, and operational boundaries used when making trade-offs | “Discuss directional ambiguity first; never `git push` without explicit authorization.” |
|
|
272
|
-
| Project knowledge | Reusable objective facts, terminology, requirements, routes, APIs, architecture, rationale, and runbooks | “A GitHub Release triggers Fuli's npm publication.” |
|
|
273
|
-
|
|
274
|
-
Four questions provide a quick classifier: “What outcome do I want?” is taste; “How do I work over
|
|
275
|
-
the long term?” is personality; “How do I decide under a trade-off?” is judgment preference; and
|
|
276
|
-
“How does this system objectively work?” is project knowledge. One-off commands, temporary output,
|
|
277
|
-
unverified guesses, raw chat, and credentials are not promoted into confirmed knowledge.
|
|
278
|
-
|
|
279
|
-
Personality is not a catch-all category for everything learned about a person. Only an explicit,
|
|
280
|
-
stable self-description can be confirmed with a human basis. A personality inferred by an Agent
|
|
281
|
-
from one behavior must remain `pending`; it cannot masquerade as human-confirmed. An empty
|
|
282
|
-
Personality filter on `/preferences` therefore usually means there is no qualifying item in the
|
|
283
|
-
selected scope, not that capture failed. “Keep copy concise” is more likely taste, while “do not
|
|
284
|
-
push without authorization” is more likely a judgment preference. To make the intent explicit,
|
|
285
|
-
say: `This is a long-term collaboration trait; save it as a personal-global personality preference: I communicate directly.` An Agent-inferred candidate can be confirmed, corrected, or invalidated
|
|
286
|
-
by the user from `/preferences` or `/flreview`.
|
|
287
|
-
|
|
288
|
-
### Personal-global, project-scoped, and shared
|
|
289
|
-
|
|
290
|
-
| Scope | When to use it | Key write fields | Retrieval rule |
|
|
291
|
-
| --- | --- | --- | --- |
|
|
292
|
-
| Personal-global preference | The behavior should remain the same in unrelated projects | `targetKind: "personal"`; omit `personalProjectId`; set `profileAspect` | Loaded for every task |
|
|
293
|
-
| Project-scoped personal preference | The collaboration behavior belongs to one project or project family | Same fields plus the exact `personalProjectId`; default to `local_only` and never inherited by child projects; explicitly use `descendants` or `selected_projects` when propagation is intended | Exact project wins; propagation follows only the explicit inheritance mode and authorized relationships, and score never resolves a conflict automatically |
|
|
294
|
-
| Project knowledge | A fact belongs to one project or to a designated parent/source project | `targetKind: "personal"`; set `personalProjectId`; omit `profileAspect` | Search the current project first, then authorized `PART_OF` / `USES_KNOWLEDGE_FROM` sources up to two hops |
|
|
295
|
-
| Team-shared knowledge | Human-confirmed project or domain knowledge genuinely needed by a team | `targetKind: "project"` after preview/review | Visible only from explicitly selected or subscribed public projects; personal preferences never enter this layer |
|
|
296
|
-
|
|
297
|
-
To choose a preference scope, ask: **Should this change the Agent's behavior in a completely
|
|
298
|
-
unrelated project?** If yes, use personal-global. If it applies only to Fuli, use a project-scoped
|
|
299
|
-
preference. If it applies to a project family, keep it project-scoped and explicitly choose whether
|
|
300
|
-
it may propagate to descendants. Do not turn an objective project fact into a personal-global item
|
|
301
|
-
merely because more than one project may reuse it. Put it in an explicit parent or knowledge-source
|
|
302
|
-
project and add a directional authorization relation. When global and project-scoped variants have
|
|
303
|
-
the same stable `attributes.preferenceKey`, human/source confirmation authority is resolved first;
|
|
304
|
-
at equal authority, the exact project variant overrides the global one.
|
|
305
|
-
|
|
306
|
-
### How Fuli identifies the current project
|
|
307
|
-
|
|
308
|
-
Fuli matches only stable `project_id` values already registered under Personal projects; it does
|
|
309
|
-
not guess from fuzzy directory similarity. Normally the `project_id` matches the repository
|
|
310
|
-
directory name, so `/workspace/fuli` maps to `fuli`. Standard calls pass only the current working
|
|
311
|
-
directory, and resolution follows this order:
|
|
312
|
-
|
|
313
|
-
1. Find the nearest Git repository root, read its `origin` remote from its `.git/config` (a
|
|
314
|
-
worktree's common directory) and match it exactly to one registered project's recorded
|
|
315
|
-
repository source URL; ssh and https forms of the same repository are equal. Any clone or
|
|
316
|
-
worktree therefore resolves to its project (`git_remote`), whatever its folder is called.
|
|
317
|
-
2. Otherwise match the repository directory name exactly to a registered `project_id`; for a
|
|
318
|
-
Codex worktree, the original repository ID from its `.git` pointer. A project that records a
|
|
319
|
-
different repository source is never matched by name (`repository_remote_conflict`).
|
|
320
|
-
3. Match the current directory name exactly to a registered project ID.
|
|
321
|
-
4. At a workspace root, match one registered direct child only when exactly one child contains
|
|
322
|
-
`.git`, `package.json`, `pyproject.toml`, `Cargo.toml`, or `go.mod`.
|
|
323
|
-
5. Multiple candidates return `ambiguous`; no candidate returns `unmatched` (a remote recorded by
|
|
324
|
-
several projects counts as several). Neither state applies project preferences or guesses a project.
|
|
325
|
-
|
|
326
|
-
For an unregistered repository with an unambiguous identity, the installed capture skill may call
|
|
327
|
-
`upsert_personal_project` before the first project-scoped write to create a minimal local private
|
|
328
|
-
project. It never creates or subscribes to a public project automatically. If a workspace has
|
|
329
|
-
multiple candidates, run the Agent from the exact project directory or explicitly register and
|
|
330
|
-
select the project first.
|
|
331
|
-
|
|
332
|
-
### Which tool the Agent calls and when
|
|
333
|
-
|
|
334
|
-
Normal use does not require manual MCP calls: submit a task to the Agent from the project directory.
|
|
335
|
-
Fuli includes lifecycle adapters for Claude Code, Codex, and Cursor. When a loaded, trusted entry
|
|
336
|
-
hook supplies `begin_task_context` context, use it without repeating the fallback. Cursor's adapter
|
|
337
|
-
starts the durable task through a hook but relies on the Agent to retrieve preferences and role
|
|
338
|
-
context. When no hook context was supplied, call the following once at task entry:
|
|
339
|
-
|
|
340
|
-
```json
|
|
341
|
-
{
|
|
342
|
-
"tool": "get_collaboration_preferences",
|
|
343
|
-
"arguments": {
|
|
344
|
-
"projectPath": "/workspace/fuli",
|
|
345
|
-
"taskPrompt": "Fix npm publishing and reuse this project's existing release conventions."
|
|
346
|
-
}
|
|
347
|
-
}
|
|
348
|
-
```
|
|
349
|
-
|
|
350
|
-
The important part of a matched response resembles:
|
|
351
|
-
|
|
352
|
-
```json
|
|
353
|
-
{
|
|
354
|
-
"context": {
|
|
355
|
-
"personal_project_id": "fuli",
|
|
356
|
-
"project_resolution": {
|
|
357
|
-
"status": "matched",
|
|
358
|
-
"basis": "repository_root",
|
|
359
|
-
"personal_project_id": "fuli"
|
|
360
|
-
}
|
|
361
|
-
},
|
|
362
|
-
"effective_preferences": ["personal-global preferences + exact fuli preferences"]
|
|
363
|
-
}
|
|
364
|
-
```
|
|
365
|
-
|
|
366
|
-
If entry-time `task_knowledge_recall` did not answer a stable project fact, search with one to four
|
|
367
|
-
short queries centered on an action, artifact, target system, or identifier. Do not use the full
|
|
368
|
-
user request verbatim as the only query:
|
|
369
|
-
|
|
370
|
-
```json
|
|
371
|
-
{
|
|
372
|
-
"tool": "search_current_project_knowledge",
|
|
373
|
-
"arguments": {
|
|
374
|
-
"projectPath": "/workspace/fuli",
|
|
375
|
-
"queries": ["npm publication runbook", "GitHub Release trigger"],
|
|
376
|
-
"includePending": false
|
|
377
|
-
}
|
|
378
|
-
}
|
|
379
|
-
```
|
|
380
|
-
|
|
381
|
-
At task end, hook mode calls `checkpoint_task_knowledge`: use `capture_candidates` only for a small,
|
|
382
|
-
durable, evidence-backed batch; otherwise use `retain_nothing`. A prompt-fallback Agent applies the
|
|
383
|
-
same rule and writes through `capture_session_knowledge`. Personal preferences set the correct
|
|
384
|
-
`profileAspect` and stable `attributes.preferenceKey`; project facts omit `profileAspect`. If
|
|
385
|
-
automatic capture is disabled under Settings, automatic writes return `capture_disabled` and no
|
|
386
|
-
category is created: transcripts, checkpoints, knowledge and preference candidates, knowledge usage,
|
|
387
|
-
view records, and feedback or corrections reported through an Agent all stop. If the capture
|
|
388
|
-
settings cannot be read, automatic writes stop the same way (`reason: policy_unavailable`). Edits
|
|
389
|
-
you submit yourself in the FULI console and an explicit `新项目:` request still save, and reading
|
|
390
|
-
existing knowledge keeps working. Session turns written while automatic recording is off are skipped
|
|
391
|
-
with a content-free marker and are never imported after it comes back on.
|
|
392
|
-
|
|
393
|
-
Capture can also be limited to projects. `projectMode` is `all` (the default), `all_except` with
|
|
394
|
-
`excludedProjectIds`, or `selected` with `selectedProjectIds`; `unregisteredDirectories`
|
|
395
|
-
(`skip` by default under `selected`, `capture` otherwise) decides work outside every registered
|
|
396
|
-
project. A directory resolves to its project the same way task entry does, so a worktree of a
|
|
397
|
-
repository is that repository's project. A project that is not recorded returns
|
|
398
|
-
`reason: project_off`, an unregistered directory `project_unregistered`, and every task receipt
|
|
399
|
-
ends with the reason and that explicit edits still save, for example
|
|
400
|
-
`此项目未开启记忆;你明确提交的修改仍会保存`. Switching a project off never hides or
|
|
401
|
-
deletes what it already recorded; `projectsOffSince` keeps when it stopped. Agents read the
|
|
402
|
-
settings with `get_capture_policy` (pass `projectPath` to learn about the current project) and
|
|
403
|
-
change them with `update_capture_policy` only after restating the change and getting the user's
|
|
404
|
-
confirmation; the settings page uses `PATCH /api/capture-policy`. Saying "这次别记" in a session pauses
|
|
405
|
-
only that session at once, without a confirmation: the Agent calls `update_capture_policy` with just
|
|
406
|
-
`sessionPaused: true` and its `taskContextToken`, and every later receipt in the session says
|
|
407
|
-
`本会话已暂停记录(自 HH:mm 起)` (`reason: session_paused`). Earlier records stay, workers the session
|
|
408
|
-
starts are paused with it, "继续记" (`sessionPaused: false`) resumes from then on, and nothing written
|
|
409
|
-
while paused is recorded later. The pause lives in the local runtime directory and ends with the session.
|
|
410
|
-
|
|
411
|
-
To turn FULI off for one session altogether, the person types `/fuli stop` (Claude Code and Cursor, where
|
|
412
|
-
setup installs the `fuli` command Skill) or `fuli stop` (any client; Codex accepts no custom slash command).
|
|
413
|
-
The prompt hook answers the line itself, so the model never sees it: from then on that session gets no
|
|
414
|
-
injected memory, preferences or recall, its Agent's own FULI reads (searches, Agent memory, conversations)
|
|
415
|
-
are refused, it records nothing and starts no collaboration (workers, loans, other Agents, role work), while other sessions, the local service and existing memories stay as they are.
|
|
416
|
-
`fuli start` turns it back on within the existing settings (a pause or a stricter global, client or project
|
|
417
|
-
setting still applies), `fuli restart` also gives the next turn the full context again, and `fuli status`
|
|
418
|
-
shows the session and the default; nothing said while FULI was off is recorded later. Context already sent
|
|
419
|
-
to the model cannot be taken back, so isolation needs a new session. Only a task token shows which session a
|
|
420
|
-
tool call is for (Claude Code keeps its tool connection across /clear; Codex and Cursor connections name no
|
|
421
|
-
session), so while any session of a client is off FULI refuses that client's tool calls without a task token,
|
|
422
|
-
rather than serve memory to a session that may be the stopped one. Agents pass the `taskContextToken` they
|
|
423
|
-
already have (entry, searches and team reads take it), so other windows keep working. The same switch runs from a shell the
|
|
424
|
-
client started (`!fuli session stop`), so an Agent with shell access can run it too (the same trust as
|
|
425
|
-
access to your files); the console and `fuli mcp --call` still read for you, and `fuli session default off` makes new sessions start with FULI
|
|
426
|
-
off until started; sessions already open keep theirs.
|
|
427
|
-
|
|
428
|
-
Run these local contract tests to check that the README and project-path behavior still match the
|
|
429
|
-
implementation:
|
|
430
|
-
|
|
431
|
-
```bash
|
|
432
|
-
npm run test:node -- test/acceptance-docs.test.js test/project-path-context.test.js
|
|
433
|
-
```
|
|
434
|
-
|
|
435
|
-
## Agent interaction sequence
|
|
436
|
-
|
|
437
|
-
This is the simplified lifecycle for a normal task. The Chinese
|
|
438
|
-
[detailed sequence diagram](acceptance/智能体调用时序图.md) also covers inheritance, usage counting,
|
|
439
|
-
write previews, and source markers.
|
|
30
|
+
<h3 align="left">Automate everything that can be automated, and hand the execution to your FLAs. You make the calls and lead the way: take your FLA crew through the hard parts, or go grab a coffee and enjoy the compound interest of your own experience.</h3>
|
|
31
|
+
|
|
32
|
+
<p align="left">Project profiles and FLA hiring run on their own if you'd rather not think about them; your only job is to set the direction. FLAs carry context and tools across every Agent client, and both top-tier and lighter models pull their weight here.</p>
|
|
33
|
+
|
|
34
|
+
|
|
35
|
+
|
|
36
|
+
Each task gets only the small slice it needs. Your judgment and experience keep adding up, so FULI
|
|
37
|
+
knows you better the more you use it. That is the compound interest. Little by little, what your
|
|
38
|
+
AI makes starts to carry your taste.
|
|
39
|
+
|
|
40
|
+
## What FULI does
|
|
41
|
+
|
|
42
|
+
<table>
|
|
43
|
+
<tr>
|
|
44
|
+
<td width="50%" valign="top">
|
|
45
|
+
<img src="docs/readme/card-memory-en.webp" alt="Claude Code, Codex and OpenCode connected to one FULI" width="460" />
|
|
46
|
+
<h3 align="center">One memory for every client</h3>
|
|
47
|
+
Connect Claude Code, Codex, OpenCode or Cursor with one click. A rule you stated in one of them is there in the others.
|
|
48
|
+
</td>
|
|
49
|
+
<td width="50%" valign="top">
|
|
50
|
+
<img src="docs/readme/card-team-en.webp" alt="Three FLAs review a fix together in the roundtable" width="460" />
|
|
51
|
+
<h3 align="center">A team of FLAs, talking at a roundtable</h3>
|
|
52
|
+
Each project gets a lead Agent (an FLA) with a name, a job and its own working memory; several FLAs form a group when the work needs it. They review, test and hand work over to each other in the roundtable, and you can read every turn.
|
|
53
|
+
</td>
|
|
54
|
+
</tr>
|
|
55
|
+
<tr>
|
|
56
|
+
<td width="50%" valign="top">
|
|
57
|
+
<img src="docs/readme/card-tonborg-en.webp" alt="Knowledge waiting to be confirmed, with Tonborg's switch" width="460" />
|
|
58
|
+
<h3 align="center"><img src="docs/readme/fla-tonborg-inline.webp" alt="" width="28" align="absmiddle" /> Tonborg confirms for you</h3>
|
|
59
|
+
New knowledge waits for a yes. Switch on Tonborg and it confirms everyday items by your preferences, asks you when it is unsure, writes down each reason, and learns from a thumbs up or down.
|
|
60
|
+
</td>
|
|
61
|
+
<td width="50%" valign="top">
|
|
62
|
+
<img src="docs/readme/card-tokens-en.webp" alt="Tokens per successful task: 0.81 times with FULI" width="460" />
|
|
63
|
+
<h3 align="center">Fewer tokens per finished task</h3>
|
|
64
|
+
Measured on 0.25.0: a successful task with FULI takes 0.81× the tokens of the same task without it, and 70 of 72 tasks pass. The saving comes from facts that live in your words, not in the code.
|
|
65
|
+
</td>
|
|
66
|
+
</tr>
|
|
67
|
+
</table>
|
|
68
|
+
|
|
69
|
+
Four FLAs come built in:
|
|
70
|
+
|
|
71
|
+
<table>
|
|
72
|
+
<tr>
|
|
73
|
+
<td width="25%" align="center" valign="top">
|
|
74
|
+
<img src="docs/readme/fla-jefa.webp" alt="Jefa" width="96" /><br />
|
|
75
|
+
<b>Jefa</b> is your project manager.
|
|
76
|
+
</td>
|
|
77
|
+
<td width="25%" align="center" valign="top">
|
|
78
|
+
<img src="docs/readme/fla-bole.webp" alt="Bole" width="96" /><br />
|
|
79
|
+
<b>Bole</b> hires FLAs and staffs each project.
|
|
80
|
+
</td>
|
|
81
|
+
<td width="25%" align="center" valign="top">
|
|
82
|
+
<img src="docs/readme/fla-tonborg.webp" alt="Tonborg" width="96" /><br />
|
|
83
|
+
<b>Tonborg</b> confirms for you and recommends clients and models.
|
|
84
|
+
</td>
|
|
85
|
+
<td width="25%" align="center" valign="top">
|
|
86
|
+
<img src="docs/readme/fla-astro.webp" alt="Astro" width="96" /><br />
|
|
87
|
+
<b>Astro</b> takes conversations that belong to no project yet.
|
|
88
|
+
</td>
|
|
89
|
+
</tr>
|
|
90
|
+
</table>
|
|
91
|
+
|
|
92
|
+
## How it works
|
|
440
93
|
|
|
441
94
|
```mermaid
|
|
442
|
-
|
|
443
|
-
|
|
444
|
-
|
|
445
|
-
|
|
446
|
-
|
|
447
|
-
|
|
448
|
-
participant G as Local graph
|
|
449
|
-
|
|
450
|
-
U->>A: Submit a task in the current project
|
|
451
|
-
alt Claude Code / Codex (hook context supplied)
|
|
452
|
-
L->>F: begin_task_context(projectPath, taskPrompt)
|
|
453
|
-
F-->>A: Preferences, exact project, bounded recall, and task token
|
|
454
|
-
else No hook context (prompt fallback, including Cursor)
|
|
455
|
-
A->>F: get_collaboration_preferences(projectPath, taskPrompt)
|
|
456
|
-
F-->>A: Preferences, exact project, and bounded recall
|
|
457
|
-
end
|
|
458
|
-
|
|
459
|
-
A->>A: Inspect task_knowledge_recall
|
|
460
|
-
opt Historical context is needed and recall did not answer it
|
|
461
|
-
A->>F: search_current_project_knowledge with focused queries
|
|
462
|
-
F->>G: Search child first, then authorized parents
|
|
463
|
-
G-->>A: Bounded content, status, provenance, and history
|
|
464
|
-
opt Knowledge materially affects the result
|
|
465
|
-
A->>F: record_knowledge_usage
|
|
466
|
-
end
|
|
467
|
-
end
|
|
468
|
-
|
|
469
|
-
A->>A: Complete the implementation, validation, or answer
|
|
470
|
-
opt The task creates a decision, contradiction, or stale signal
|
|
471
|
-
A->>F: Record rationale or knowledge feedback
|
|
472
|
-
end
|
|
473
|
-
opt Something is worth keeping (candidates, a decision, a meaningful work log)
|
|
474
|
-
A->>F: checkpoint_task_knowledge
|
|
475
|
-
F->>G: capture_candidates
|
|
476
|
-
end
|
|
477
|
-
|
|
478
|
-
alt Loaded and trusted Claude Code / Codex Stop hook
|
|
479
|
-
L->>F: verify_task_checkpoint
|
|
480
|
-
F-->>L: Close a routine task as retain_nothing / reported, otherwise remind
|
|
481
|
-
else Cursor or an Agent without a Stop hook
|
|
482
|
-
Note over A,F: Same contract. A routine fallback task closes at the next task entry
|
|
483
|
-
end
|
|
484
|
-
A-->>U: Return the result and any Fuli sources actually used
|
|
95
|
+
flowchart LR
|
|
96
|
+
U[You] -->|ask inside a project| A[Agent client]
|
|
97
|
+
A -->|task starts| F[FULI on your computer]
|
|
98
|
+
F -->|preferences and a small recall| A
|
|
99
|
+
A -->|look up more if needed| F
|
|
100
|
+
A -->|write back what is reusable| F
|
|
485
101
|
```
|
|
486
102
|
|
|
487
|
-
|
|
488
|
-
|
|
489
|
-
One knowledge-base connection can be bound to one or more existing personal projects without
|
|
490
|
-
granting Fuli write access to the source. Each target keeps independent retrieval mode, sync
|
|
491
|
-
cursor, and error state. The Connections page supports multi-project assignment, connection
|
|
492
|
-
checks, manual synchronization, disconnection, and a per-project conflict switch.
|
|
493
|
-
|
|
494
|
-
| Connector | Read path | Current boundary |
|
|
495
|
-
| --- | --- | --- |
|
|
496
|
-
| MCP | Resources `list/read`; configured `search` / `fetch` tools | Default; HTTPS or loopback HTTP, plus stdio |
|
|
497
|
-
| Notion | Page Markdown and Data Source queries | Current `2026-03-11` API; explicit page or data-source IDs |
|
|
498
|
-
| Feishu / Lark | Wiki nodes, search, and Docx raw content | Docx text only; search may require a user access token |
|
|
499
|
-
| RAG Retrieval API | Dify-compatible external-knowledge retrieval contract | Live only; for compatible endpoints or adapters in front of systems such as RAGFlow |
|
|
500
|
-
| Trusted custom code | Local ESM `sync` / `retrieve` contract | Explicitly installed trusted code, not a sandbox |
|
|
501
|
-
|
|
502
|
-
<a id="connect-external-knowledge"></a>
|
|
503
|
-
|
|
504
|
-
### Connect one or more knowledge bases to a project
|
|
505
|
-
|
|
506
|
-
One personal project can have multiple knowledge connections, and one connection can target
|
|
507
|
-
multiple personal projects. Each submission adds a connection instead of replacing an existing
|
|
508
|
-
one. Source configuration is stored once, while retrieval mode, synchronization state, and errors
|
|
509
|
-
are isolated per target. Use Bind projects on an existing connection to add or remove targets.
|
|
510
|
-
|
|
511
|
-
1. Create or confirm the target under Personal projects.
|
|
512
|
-
2. If the source needs a token, set it in the environment that starts Fuli, then start or restart
|
|
513
|
-
the service. The binding stores only the environment-variable name. For example, run
|
|
514
|
-
`export PROJECT_KB_TOKEN='...'` and then `fl restart`.
|
|
515
|
-
3. Open `http://127.0.0.1:2727/connections`. Under External knowledge, enter a connection name,
|
|
516
|
-
connector, one or more bound projects, and retrieval mode.
|
|
517
|
-
4. For MCP HTTP, enter the final endpoint URL (redirects to another origin are not followed) and an
|
|
518
|
-
optional token environment-variable name. For MCP stdio,
|
|
519
|
-
enter its command and arguments. For Notion, enter the token environment-variable name and Page
|
|
520
|
-
IDs or Data Source IDs. For Feishu/Lark, enter the token environment-variable name, region, and
|
|
521
|
-
an explicit Space ID, root-node token, or node tokens. For RAG Retrieval API, enter a
|
|
522
|
-
Dify-compatible endpoint, knowledge-base IDs, and an optional token environment-variable name.
|
|
523
|
-
A custom connector takes a trusted local ESM module and source JSON.
|
|
524
|
-
5. Select Add connection and then Check. A successful `mirror` or `hybrid` binding can be synced;
|
|
525
|
-
a `live` binding reads its source when an Agent query runs.
|
|
526
|
-
6. After creation, use Bind projects to add or remove projects and select a mode per project.
|
|
527
|
-
|
|
528
|
-
The project graph shows an External knowledge source node and a Uses external knowledge edge for
|
|
529
|
-
each assignment. This is a read-only projection of connection configuration. Live bodies are still
|
|
530
|
-
retrieved only when an Agent invokes `search_connected_knowledge`; showing the node does not mirror
|
|
531
|
-
content locally.
|
|
532
|
-
|
|
533
|
-
Retrieval modes differ as follows:
|
|
534
|
-
|
|
535
|
-
- `live`: read the third party at query time without storing a body mirror in Fuli; the source must
|
|
536
|
-
be available.
|
|
537
|
-
- `mirror`: manually mirror read-only content into the bound personal project and query the graph.
|
|
538
|
-
- `hybrid`: use both the local mirror and live source results; recommended when the connector
|
|
539
|
-
supports both capabilities.
|
|
540
|
-
|
|
541
|
-
Bindings support `live`, `mirror`, and `hybrid` modes. Mirror content is written only to the bound
|
|
542
|
-
personal project as `observed`, `pending`, `restricted`, local-only knowledge. Credentials are read
|
|
543
|
-
from named environment variables and are not stored in binding JSON. Source writes are never part
|
|
544
|
-
of the connector contract. Documents that look like credential material are skipped individually,
|
|
545
|
-
reported by count, and never enter the graph or Agent context.
|
|
546
|
-
|
|
547
|
-
<a id="external-knowledge-conflict-policy"></a>
|
|
103
|
+
A task starts with the preferences in effect and a few related facts. The Agent looks up the rest
|
|
104
|
+
when it needs it, and at the end writes back only what the next task can reuse.
|
|
548
105
|
|
|
549
|
-
|
|
106
|
+
## Install
|
|
550
107
|
|
|
551
|
-
|
|
552
|
-
(Connections → Conflict policy). The policy covers that project's personal graph, all external
|
|
553
|
-
bindings, and explicitly selected public projects:
|
|
108
|
+
FULI runs a small service and a console on your own computer.
|
|
554
109
|
|
|
555
|
-
|
|
556
|
-
source, and scope separately so the user can choose.
|
|
557
|
-
- Allow an Agent decision lets the Agent select the stronger or fresher source for the current
|
|
558
|
-
response and requires it to explain the basis. It never confirms, invalidates, overwrites, or
|
|
559
|
-
writes back any knowledge.
|
|
110
|
+
**Desktop client (recommended).** Download from the [latest release](https://github.com/JevonsCode/fuli/releases/latest):
|
|
560
111
|
|
|
561
|
-
|
|
562
|
-
|
|
563
|
-
|
|
564
|
-
|
|
565
|
-
Agent finds a material conflict, the default policy is to show it in the conversation and ask the
|
|
566
|
-
user; the optional Agent-decision policy applies to the current response only and cannot confirm,
|
|
567
|
-
invalidate, or rewrite any source. Direct external-to-public binding, reviewed public promotion,
|
|
568
|
-
scheduled sync, webhooks, and full deletion reconciliation remain **TODO**.
|
|
569
|
-
|
|
570
|
-
See the [external-knowledge architecture and connector contract](docs/external-knowledge-architecture.md), and the [public-workspace and personal-graph architecture](docs/public-personal-architecture.md).
|
|
571
|
-
|
|
572
|
-
## Current capabilities and evidence boundaries
|
|
573
|
-
|
|
574
|
-
Core mechanisms covered by implementation and automated tests include:
|
|
575
|
-
|
|
576
|
-
- a local personal space, exact project scope, and selective parent inheritance;
|
|
577
|
-
- preference confirmation, deferred conflict resolution, revision history, and source markers;
|
|
578
|
-
- bounded task-prompt recall, focused on-demand retrieval, and material-usage auditing;
|
|
579
|
-
- task entry and completion checkpoints, plus local Claude Code, Codex, and Cursor lifecycle adapters;
|
|
580
|
-
- decision options, rationale, and validation results supplied during initial capture;
|
|
581
|
-
- knowledge-usage events, negative feedback, and content-generation isolation;
|
|
582
|
-
- persistent scoped knowledge reviews with pause/resume watermarks;
|
|
583
|
-
- common-knowledge discovery and preview-token-protected atomic promotion;
|
|
584
|
-
- read-only multi-project MCP, Notion, Feishu/Lark, Dify-compatible RAG Retrieval API, and trusted-code bindings with connected retrieval.
|
|
585
|
-
|
|
586
|
-
New in 0.20 (public preview `0.20.0-beta`, see the [release notes](docs/releases/0.20.0.md)):
|
|
587
|
-
|
|
588
|
-
- **Projects and people**: unassigned conversations go to Astro, Jefa triages them, and when a topic keeps coming back (three conversations over at least two days) Astro asks whether to make it a project with its own people. Every project has a brief and every Agent an occupation, which Bole recruits by.
|
|
589
|
-
- **Tasks**: assign without starting (backlog), and stop, retry, or review any run from the task details. A temporary failure is retried automatically only when no prompt was sent.
|
|
590
|
-
- **Decisions**: an Agent that needs you asks in the conversation first. A request that must wait for you goes to the Decisions page, and your reply reaches the Agent that asked.
|
|
591
|
-
- **Roundtable**: besides asks (`message_agent`), Agents have direct and group chats (`agent_chat`) with push delivery and `wait`; the console shows them live.
|
|
592
|
-
- **Memory between projects**: weighted links (related, frequent, close) say how much one project may draw on another. Weights never grant access.
|
|
593
|
-
- **Lower token cost**: a smaller task entry, shared rules read on demand, long preference lists read in pages, and a Stop hook that closes routine tasks without another model turn.
|
|
594
|
-
- **Console**: six entries (Workbench, Agents, Conversations, Projects, Memory, Settings), a phone layout, and Alt/Option-click explanations for any status or term.
|
|
595
|
-
- **Windows**: download `FULI-Setup-<version>.exe` from the release and double-click it to install for your user (no terminal, no administrator; unsigned for now, so SmartScreen asks once). The installed build has Update-Fuli, Rollback-Fuli, and Uninstall-Fuli that keeps your data; the console's 「Update to vX」 downloads the release's verified Windows package and runs the same update in one click.
|
|
596
|
-
|
|
597
|
-
The following must not be presented as already proven:
|
|
598
|
-
|
|
599
|
-
- the A/B thresholds in `FULI_ALIGNMENT_BENCHMARK.md` are acceptance conventions; claims that Fuli
|
|
600
|
-
reduces explanation or rework require a sufficiently large set of controlled, real paired tasks;
|
|
601
|
-
- benchmark projects and conversations are explicitly labeled **MOCK / synthetic data**, not user or
|
|
602
|
-
production data;
|
|
603
|
-
- the decision tool can include validation during initial capture, but it does not yet expose a
|
|
604
|
-
dedicated operation for appending immutable validation results to an existing `Decision`;
|
|
605
|
-
- Installed files do not prove host loading, trust, or execution. Claude Code and Codex adapters
|
|
606
|
-
generate entry/Stop hook configuration; Codex trust review remains a user action. Cursor uses
|
|
607
|
-
task-entry hooks and bounded Stop reminders, not an equivalent unconditional Stop gate. Ordinary
|
|
608
|
-
input in each real client still requires live acceptance;
|
|
609
|
-
- selected public projects in connected retrieval are Beta; external-to-public promotion and
|
|
610
|
-
background delta synchronization are not implemented.
|
|
611
|
-
|
|
612
|
-
## Installation
|
|
613
|
-
|
|
614
|
-
Requirements:
|
|
112
|
+
| System | File |
|
|
113
|
+
| --- | --- |
|
|
114
|
+
| Windows | `FULI-Setup-<version>.exe`. No terminal and no administrator rights. |
|
|
115
|
+
| macOS (Apple silicon, 13.5 or later) | `fuli-macos-arm64-<version>.dmg`. Drag FULI onto Applications. |
|
|
615
116
|
|
|
616
|
-
|
|
617
|
-
|
|
618
|
-
|
|
619
|
-
- on a memory-constrained Mac, use
|
|
620
|
-
`fuli setup --runtime-mode native --memory-profile low --adaptive-memory`. Container mode remains the compatibility default and is never changed automatically.
|
|
117
|
+
On first launch FULI finds the AI clients on this computer and connects each with one click. The
|
|
118
|
+
installers are not signed yet, so each system holds FULI back once; [Install and run](docs/install.md)
|
|
119
|
+
shows the two clicks that let it through.
|
|
621
120
|
|
|
622
|
-
|
|
121
|
+
**Servers, VMs and the cloud.** Without a desktop, and on Intel Macs and Linux, install with npm
|
|
122
|
+
(Node.js 24.12 or later):
|
|
623
123
|
|
|
624
124
|
```bash
|
|
625
125
|
npm install --global fuli-context
|
|
626
126
|
fuli setup
|
|
627
|
-
```
|
|
628
|
-
|
|
629
|
-
Or use the one-line installer, which checks the Node.js version first:
|
|
630
|
-
|
|
631
|
-
```bash
|
|
632
|
-
curl -fsSL https://xn--8ovp9s.xn--m8txu.com/fuli/install.sh | sh
|
|
633
|
-
```
|
|
634
|
-
|
|
635
|
-
On Windows PowerShell: `irm https://xn--8ovp9s.xn--m8txu.com/fuli/install.ps1 | iex`.
|
|
636
|
-
|
|
637
|
-
`fuli setup` first shows its plan and asks for confirmation. It then checks the container runtime,
|
|
638
|
-
initializes local Graphiti / Neo4j, creates a personal space, installs the companion Agent Skills,
|
|
639
|
-
and registers the `fuli` MCP with detected agents. The default setup connects only the personal
|
|
640
|
-
Provider; it does not simulate a team-shared service.
|
|
641
|
-
|
|
642
|
-
Open the management UI after setup:
|
|
643
|
-
|
|
644
|
-
```bash
|
|
645
127
|
fuli open
|
|
646
128
|
```
|
|
647
129
|
|
|
648
|
-
|
|
649
|
-
between boots (`netsh interface ipv4 show excludedportrange protocol=tcp` lists them). A first setup picks ports outside
|
|
650
|
-
those ranges. Every `fuli start`, `restart` and `setup` checks again: a Fuli port that is now reserved, or a console port
|
|
651
|
-
Windows refuses or another program holds, moves to the nearest free port. Fuli backs up `runtime-settings.json`, saves
|
|
652
|
-
the new port, starts normally and says so once (in the command output, the desktop client and the console). It never
|
|
653
|
-
moves away from another running Fuli console on that port, and never starts a second one beside it. The Windows tray's
|
|
654
|
-
**Use another port** still moves the console to a free port near the default on request.
|
|
655
|
-
Open Agent sessions keep showing the old console address in Fuli links until they restart.
|
|
656
|
-
|
|
657
|
-
The Settings page manages all seven local Fuli ports, graph runtime mode, LAN access, automatic capture,
|
|
658
|
-
Agent access, UI language, and the resource refresh interval. Run `fuli restart` after saving port,
|
|
659
|
-
runtime, or LAN changes;
|
|
660
|
-
the refresh interval applies immediately. The same page polls real memory and disk usage for the
|
|
661
|
-
management service, Providers, Neo4j, application files, and local data. Memory is sampled on every
|
|
662
|
-
poll; disk usage is sampled at most once per minute, with separate timestamps. Missing container metrics
|
|
663
|
-
are shown as partial instead of being replaced with mock values. Native Provider and Neo4j processes
|
|
664
|
-
are included. Browser-tab memory is excluded; shared container-VM overhead is excluded in container mode.
|
|
665
|
-
|
|
666
|
-
## CLI
|
|
667
|
-
|
|
668
|
-
The global package provides two equivalent commands: `fuli` and the short alias `fl`.
|
|
669
|
-
CLI help, prompts, status output, and errors are always in English and do not follow the web or
|
|
670
|
-
system locale. User data such as paths and space names is displayed unchanged.
|
|
671
|
-
|
|
672
|
-
| Command | Purpose |
|
|
673
|
-
| --- | --- |
|
|
674
|
-
| `fuli --help` / `fuli -h` | List every public command and its available options |
|
|
675
|
-
| `fuli --version` / `fuli -v` | Print the installed Fuli CLI version |
|
|
676
|
-
| `fuli setup [options]` | Initialize the Provider, UI, agent integrations, and Skills; safe to rerun |
|
|
677
|
-
| `fuli start [options]` | Check Agent integrations, then start the Provider and UI; optionally open a browser, rebuild containers, or enable LAN access |
|
|
678
|
-
| `fuli stop [--data-dir DIR]` | Stop services without deleting data |
|
|
679
|
-
| `fuli restart [options]` | Restart local services using the same runtime options as `start` |
|
|
680
|
-
| `fuli status [--json] [--data-dir DIR] [--port PORT]` | Show UI, personal graph, and shared-service status; `--json` emits machine-readable output |
|
|
681
|
-
| `fuli open [--data-dir DIR]` | Open the current management UI in the default browser |
|
|
682
|
-
| `fuli graph export --output DIR [--mode container\|native]` | Export graph data as a checksummed, copyable offline bundle |
|
|
683
|
-
| `fuli graph import --input DIR [--target-mode container\|native] [--yes]` | Verify and replace target graph data, preserving an automatic pre-import rollback bundle |
|
|
684
|
-
| `fuli graph migrate --to native [--yes]` | Move a container graph to native mode, accepted only when node and relationship counts match |
|
|
685
|
-
| `fuli update [setup options]` | Update the npm package and refresh local integrations |
|
|
686
|
-
| `fuli uninstall [--yes] [--data-dir DIR]` | Remove agent integrations and services while preserving knowledge data and Neo4j volumes |
|
|
687
|
-
|
|
688
|
-
Common commands:
|
|
689
|
-
|
|
690
|
-
```bash
|
|
691
|
-
fuli --version
|
|
692
|
-
fuli status
|
|
693
|
-
fuli restart --rebuild
|
|
694
|
-
fuli start --lan
|
|
695
|
-
fuli graph export --output "$HOME/Backups/fuli-graph"
|
|
696
|
-
fuli stop
|
|
697
|
-
```
|
|
698
|
-
|
|
699
|
-
Connect a deployed `fuli-workspace` service:
|
|
700
|
-
|
|
701
|
-
```bash
|
|
702
|
-
fuli connect-workspace \
|
|
703
|
-
--url http://127.0.0.1:8789 \
|
|
704
|
-
--token-file /path/to/private-token
|
|
705
|
-
fuli restart
|
|
706
|
-
```
|
|
707
|
-
|
|
708
|
-
The command verifies discovery, protocol version, and token scope before atomically updating the
|
|
709
|
-
local runtime configuration with mode `0600`. It never prints the token, token-file path, or remote
|
|
710
|
-
principal ID. Non-loopback services must use HTTPS. The current `fuli-workspace-v1` adapter enables
|
|
711
|
-
discovery, explicit subscriptions, and queries; publishing, contributions, and review remain off
|
|
712
|
-
until those protocol mappings are implemented.
|
|
713
|
-
|
|
714
|
-
`start` and `restart` accept `--data-dir DIR`, `--personal-space NAME`, `--port PORT`, `--open`,
|
|
715
|
-
`--rebuild`, `--lan`, and `--no-lan`. When port or LAN flags are omitted, the saved Settings values
|
|
716
|
-
are used. LAN mode listens on private IPv4 addresses and prints the reachable URLs, the
|
|
717
|
-
`fuli` username, and a temporary access code generated for that start. The default remains
|
|
718
|
-
loopback-only; internal Providers and Neo4j ports are never exposed by `--lan`. LAN mode uses
|
|
719
|
-
HTTP Basic Auth and is intended only for trusted home or office Wi-Fi, not public deployment.
|
|
720
|
-
Before starting, `fuli start` performs a read-only check of detected Agent MCP, Skill, Codex bootstrap,
|
|
721
|
-
and Claude lifecycle integrations. If any are missing or outdated, it still starts the local services
|
|
722
|
-
and tells you to run `fuli setup`; it never applies setup changes implicitly.
|
|
723
|
-
|
|
724
|
-
`setup` and `update` accept:
|
|
725
|
-
|
|
726
|
-
| Option | Purpose |
|
|
727
|
-
| --- | --- |
|
|
728
|
-
| `--yes` | Skip confirmation for unattended setup |
|
|
729
|
-
| `--codex-only` | Configure Codex only |
|
|
730
|
-
| `--data-dir DIR` | Use a specific data and configuration directory |
|
|
731
|
-
| `--personal-space NAME` | Set the personal-space name; defaults to `Personal` |
|
|
732
|
-
| `--port PORT` | Set the management UI port; defaults to `2727` |
|
|
733
|
-
| `--runtime-mode container\|native` | Select containers or native processes; defaults to `container`, with native mode currently on macOS/Linux |
|
|
734
|
-
| `--memory-profile low\|balanced` | Select Neo4j's memory budget; fresh installs default to `balanced` |
|
|
735
|
-
| `--adaptive-memory` | Enable on-demand wake and staged idle sleep, coordinated by the management service |
|
|
736
|
-
| `--no-adaptive-memory` | Disable idle sleep and keep graph services running |
|
|
737
|
-
| `--skip-agents` | Do not change agent configuration or Skills |
|
|
738
|
-
| `--connected-only` | Refresh only clients that already have a Fuli MCP entry; used automatically by update |
|
|
739
|
-
| `--no-start` | Initialize the Provider without starting the UI |
|
|
740
|
-
| `--allow-provider-downgrade` | Accept a lower native Provider version, for example moving from a local `rc` build to a public `beta`; back up the data folder first |
|
|
741
|
-
| `--personal-only` | Use only the personal Provider; this is the default |
|
|
742
|
-
| `--with-dev-public` | Start a development shared Provider for local integration work only |
|
|
743
|
-
|
|
744
|
-
The `low` profile uses a 128 MiB initial heap, 256 MiB maximum heap, and 64 MiB page cache.
|
|
745
|
-
`balanced` uses 256 MiB, 512 MiB, and 256 MiB respectively. The low profile does not add a hard
|
|
746
|
-
container memory limit: it lowers steady memory use while leaving room for Neo4j's native memory.
|
|
747
|
-
It may cause more garbage collection and disk reads under bulk writes, large traversals, or high
|
|
748
|
-
concurrency, but it does not change transaction semantics, storage format, or data volumes. If those
|
|
749
|
-
workloads become slow or fail for lack of heap, run `fuli setup --memory-profile balanced --yes`.
|
|
750
|
-
The selected profile is saved and reused by later setup, update, start, and restart commands.
|
|
751
|
-
|
|
752
|
-
Adaptive memory is a lifecycle policy independent of the Neo4j memory profile. When enabled, the
|
|
753
|
-
lightweight management service stays available. Real MCP tools, MCP resources, and graph requests
|
|
754
|
-
acquire a runtime lease and wake services on demand. After the final lease is released, the personal
|
|
755
|
-
Provider stops after 60 idle seconds by default and Neo4j stops after 180 idle seconds. Volumes are
|
|
756
|
-
preserved and the next request restores the same data. Active calls renew their leases and cannot be
|
|
757
|
-
interrupted by idle timers; a crashed client lease expires after at most 180 seconds. These durations
|
|
758
|
-
are current product defaults, not measured hardware thresholds.
|
|
759
|
-
|
|
760
|
-
This mode trades idle memory for first-request cold-start latency and additional disk reads. Health,
|
|
761
|
-
runtime-status, and resource polling do not wake the graph, and `fuli status` treats intentional sleep
|
|
762
|
-
as healthy. The management service must remain running to wake graph services, so `--no-start` only
|
|
763
|
-
saves the policy for a later start. In container mode the policy stops only Fuli Provider and Neo4j
|
|
764
|
-
containers; it does not shut down Rancher Desktop, Docker Desktop, Kubernetes, or the container VM
|
|
765
|
-
itself. Native mode directly stops the corresponding Provider and Neo4j processes, so no shared VM
|
|
766
|
-
overhead remains while idle.
|
|
767
|
-
|
|
768
|
-
Jefa (project manager), Bole (AR, Agent Resources), and Tonborg (judgment) are built-in roles in every
|
|
769
|
-
space and are never hired. Bole's workbench shows staffing, current work, and recruitment history;
|
|
770
|
-
Tonborg's workbench shows decisions and feedback. Jefa's task board is built into FULI and needs no
|
|
771
|
-
installation; **Manage assigned projects** on a role's card or workbench sets which projects it manages.
|
|
772
|
-
The console has no hiring step: a custom employee package is recruited through the `recruit_employee` Agent
|
|
773
|
-
tool, which reuses a durable identity with separate project assignments. An installed employee workbench and
|
|
774
|
-
API share the FULI port; existing FULI MCP clients discover and call its tools. Recruitment does not start a
|
|
775
|
-
model or rewrite client settings. See the [employee package and extension contract](docs/employee-agents.md).
|
|
776
|
-
Agents can also grow small [Agent apps](docs/agent-apps.md); in this version an approved app runs
|
|
777
|
-
unsandboxed with your account's permissions, after you confirm it with `fl app approve` in a terminal.
|
|
778
|
-
|
|
779
|
-
Project Agent identities remain control-plane records rather than one resident process per identity.
|
|
780
|
-
Roles now keep versioned, project-private working memory in the same Neo4j Provider. Task entry
|
|
781
|
-
restores one owner across connected MCP clients; lifecycle checkpoints survive separate
|
|
782
|
-
MCP processes. See [role memory, host hooks and acceptance boundaries](docs/project-agent-memory.md).
|
|
783
|
-
Supported Codex and Claude Code hooks also journal visible conversations. Once unfolded messages pass
|
|
784
|
-
a size threshold (64 KB by default), older ones fold into a digest; recovery stays within a byte budget. Specialist loans and artifact verification share the same Provider.
|
|
785
|
-
See [conversation continuity and collaboration](docs/agent-conversations-and-collaboration.md)
|
|
786
|
-
for client coverage, model capability configuration and acceptance limits.
|
|
787
|
-
Physical executors share leases by executor ID. Fuli starts and stops only executors with an explicitly
|
|
788
|
-
injected managed lifecycle adapter; host-owned external executors such as Codex are never started or
|
|
789
|
-
killed implicitly. The current minimum-memory combination is:
|
|
790
|
-
|
|
791
|
-
```bash
|
|
792
|
-
fuli setup --yes --runtime-mode native --memory-profile low --adaptive-memory
|
|
793
|
-
```
|
|
794
|
-
|
|
795
|
-
The two modes use separate data directories. Switching preserves the old mode's data and never
|
|
796
|
-
merges changes in the background. To migrate, export while the source mode is still active, then
|
|
797
|
-
install or switch to the target mode and import:
|
|
130
|
+
`fuli setup` shows its plan and changes nothing until you confirm.
|
|
798
131
|
|
|
799
|
-
|
|
800
|
-
# Rancher / Docker -> native
|
|
801
|
-
fuli graph export --mode container --output "$HOME/Backups/fuli-container"
|
|
802
|
-
fuli setup --yes --runtime-mode native --memory-profile low --adaptive-memory
|
|
803
|
-
fuli graph import --target-mode native --input "$HOME/Backups/fuli-container" --yes
|
|
804
|
-
```
|
|
805
|
-
|
|
806
|
-
Or run the same steps as one command, from a package that has the native runtime (on Windows the
|
|
807
|
-
installed Fuli desktop package):
|
|
808
|
-
|
|
809
|
-
```bash
|
|
810
|
-
fuli graph migrate --to native
|
|
811
|
-
```
|
|
812
|
-
|
|
813
|
-
It stops the running console, counts and exports the container graph, sets up native mode, imports,
|
|
814
|
-
and accepts the result only when every label and relationship-type count matches. On any failure it
|
|
815
|
-
switches back to container mode. The container volume is only read and is left stopped, never removed.
|
|
816
|
-
Client integrations are not changed.
|
|
132
|
+
## The console
|
|
817
133
|
|
|
818
|
-
|
|
819
|
-
|
|
820
|
-
|
|
821
|
-
inside the data directory and attempts an immediate rollback if loading fails. Let active Agent graph
|
|
822
|
-
writes finish before either operation.
|
|
134
|
+
<p align="center">
|
|
135
|
+
<img src="docs/readme/console-en.webp" alt="The FULI console: a task in review, a question waiting for you and the FLAs at work" width="800" />
|
|
136
|
+
</p>
|
|
823
137
|
|
|
824
|
-
|
|
825
|
-
|
|
826
|
-
to the bundle. Unmanaged external Workspace connections are preserved for a same-device migration.
|
|
827
|
-
Reconnect their credentials separately after moving the bundle to another device.
|
|
138
|
+
The console opens on what matters now: questions waiting for you, the FLAs at work and what they
|
|
139
|
+
finished. Everything runs on `127.0.0.1`; there is no account and no sign-in.
|
|
828
140
|
|
|
829
|
-
|
|
830
|
-
`workspace.dump`; it can be copied to another disk or machine. Container mode transfers files through
|
|
831
|
-
the Docker API, so the backup location does not need to be shared with Rancher. Dumps use the pinned
|
|
832
|
-
Neo4j 5.26 format and are intended for the two Fuli runtime modes or a compatible Neo4j 5.26 setup;
|
|
833
|
-
they are not a generic CSV/JSON interchange format for arbitrary databases.
|
|
141
|
+
## FAQ
|
|
834
142
|
|
|
835
|
-
|
|
143
|
+
<details>
|
|
144
|
+
<summary><b>Where is my data?</b></summary>
|
|
836
145
|
|
|
837
|
-
|
|
838
|
-
|
|
839
|
-
|
|
146
|
+
On your computer. Your personal graph and conversation records stay local. External knowledge
|
|
147
|
+
(Notion, Lark, MCP or RAG sources) is read-only: FULI never writes back.
|
|
148
|
+
</details>
|
|
840
149
|
|
|
841
|
-
|
|
842
|
-
|
|
843
|
-
it verifies that it will not downgrade, stops the old services, installs the latest version, and
|
|
844
|
-
refreshes only clients with an existing Fuli MCP entry. Existing knowledge, configuration backups,
|
|
845
|
-
and Neo4j volumes remain intact. Reuse any custom setup options during updates.
|
|
150
|
+
<details>
|
|
151
|
+
<summary><b>Which clients work with FULI?</b></summary>
|
|
846
152
|
|
|
847
|
-
|
|
848
|
-
|
|
849
|
-
|
|
850
|
-
(no Fuli MCP entry) require a full `fuli setup`.
|
|
153
|
+
Claude Code, Codex, OpenCode, Cursor and Claude Desktop connect from the console with one click.
|
|
154
|
+
Other MCP clients run `fuli connect` for a standard configuration.
|
|
155
|
+
</details>
|
|
851
156
|
|
|
852
|
-
|
|
157
|
+
<details>
|
|
158
|
+
<summary><b>How do I call one of my FLAs?</b></summary>
|
|
853
159
|
|
|
854
|
-
|
|
855
|
-
|
|
856
|
-
|
|
857
|
-
```
|
|
160
|
+
Start a message with `#FLA` and its name. In Claude Code, OpenCode and Cursor you can also type
|
|
161
|
+
`/fla` and pick one from the menu; in the Codex app, type `/prompts:fla`. Turn the menu off in Settings › Clients.
|
|
162
|
+
</details>
|
|
858
163
|
|
|
859
|
-
|
|
860
|
-
|
|
164
|
+
<details>
|
|
165
|
+
<summary><b>How does FULI save tokens?</b></summary>
|
|
861
166
|
|
|
862
|
-
|
|
167
|
+
A session starts with only the preferences in effect and a few related facts; the Agent looks up
|
|
168
|
+
the rest when it needs it. Facts that only you know, such as a sandbox URL you mentioned once,
|
|
169
|
+
come back later without a search through the code. The numbers are in the
|
|
170
|
+
[benchmark](FULI_ALIGNMENT_BENCHMARK.md) (section 8.9.4).
|
|
171
|
+
</details>
|
|
863
172
|
|
|
864
|
-
|
|
865
|
-
|
|
866
|
-
fresh interactive review artifact. Codex rejects custom slash commands before any hook runs, so there the person
|
|
867
|
-
types `fuli review` (or `flreview`) on its own line; the Codex prompt hook passes that exact line to the model with
|
|
868
|
-
the installed Skill named, and in a session where FULI is off it answers the line itself without starting a review. FULI remains the durable source of truth and write surface; no
|
|
869
|
-
permanent review tab or one-question-at-a-time chat flow is required.
|
|
870
|
-
`operating-fuli` makes Agents inspect the structured interface-parity catalog before changing data,
|
|
871
|
-
verify writes through matching read tools, and treat browser automation as visual verification only.
|
|
872
|
-
See the [Agent interface architecture](docs/agent-interface-architecture.md).
|
|
873
|
-
Fuli setup generates Claude Code `UserPromptSubmit`/`Stop` MCP hooks, a Codex
|
|
874
|
-
`UserPromptSubmit` MCP hook plus a local `Stop` command, and Cursor
|
|
875
|
-
`sessionStart`/`beforeSubmitPrompt`/`stop` hooks. The Codex Stop command first blocks an unfinished
|
|
876
|
-
checkpoint; if the same token is still unfinished after one continuation, it records only
|
|
877
|
-
`retain_nothing` to prevent an infinite loop and never invents knowledge or Agent memory. Codex
|
|
878
|
-
bootstrap and Cursor instructions retain the fallback for tasks without supplied lifecycle context;
|
|
879
|
-
they must not duplicate supplied preferences. Installing configuration never approves host trust.
|
|
880
|
-
Preference content remains in local Fuli as the single source of truth and is not copied into agent
|
|
881
|
-
configuration.
|
|
173
|
+
<details>
|
|
174
|
+
<summary><b>Do the built-in FLAs cost extra?</b></summary>
|
|
882
175
|
|
|
883
|
-
|
|
884
|
-
|
|
885
|
-
|
|
176
|
+
By default the built-in FLAs work inside your client's session. Only when you connect a model for
|
|
177
|
+
them do they use their own model, so nothing spends money behind your back.
|
|
178
|
+
</details>
|
|
886
179
|
|
|
887
|
-
|
|
180
|
+
<details>
|
|
181
|
+
<summary><b>How do I take FULI out of a client?</b></summary>
|
|
888
182
|
|
|
889
|
-
|
|
890
|
-
|
|
891
|
-
|
|
183
|
+
In the console, open Settings › Clients and choose Remove for that client. `fuli uninstall` and the
|
|
184
|
+
Windows uninstaller do the same for every connected client. FULI takes out only its own entries
|
|
185
|
+
(the MCP server, hooks, Skills and its permission lines), backs up each file first, and keeps a
|
|
186
|
+
Skill you edited yourself. Your FULI data stays until you delete it. On macOS, remove the clients
|
|
187
|
+
in the console before you move FULI.app to the Trash: the Trash alone leaves those entries in place.
|
|
188
|
+
</details>
|
|
892
189
|
|
|
893
|
-
|
|
894
|
-
|
|
895
|
-
openssl rand -hex 32 > ./fuli-remote.token
|
|
896
|
-
fuli remote-mcp --personal-project-id YOUR_PROJECT_ID \
|
|
897
|
-
--bearer-token-file ./fuli-remote.token --port 2728
|
|
898
|
-
```
|
|
190
|
+
<details>
|
|
191
|
+
<summary><b>I installed FULI with npm before. Is it installed twice now?</b></summary>
|
|
899
192
|
|
|
900
|
-
|
|
901
|
-
|
|
902
|
-
|
|
193
|
+
The program is, your data is not: both use the same data folder. When the desktop client starts, it runs the one FULI
|
|
194
|
+
service and points your connected clients at itself. The console then offers to remove the npm copy with one click;
|
|
195
|
+
until it is removed, a bare `fl` in a terminal is still the npm version.
|
|
196
|
+
[Details](docs/install.md#already-installed-with-npm)
|
|
197
|
+
</details>
|
|
903
198
|
|
|
904
|
-
|
|
905
|
-
|
|
906
|
-
with `--max-sessions` and `--session-idle-ttl-seconds` when needed. Idle means no completed MCP
|
|
907
|
-
application request: a silent standalone SSE notification stream does not pin a slot forever and
|
|
908
|
-
may be closed at the TTL, after which the client must reconnect and initialize a new session.
|
|
909
|
-
Shutdown gives bounded requests and session cleanup one shared grace deadline before forcing the
|
|
910
|
-
remaining HTTP connections closed. When a reverse proxy supplies a
|
|
911
|
-
browser Origin or preserves its public Host, add the exact values with
|
|
912
|
-
`--allowed-origin https://connector.example --allowed-host connector.example`. Otherwise configure
|
|
913
|
-
the proxy to rewrite Host to the loopback upstream authority. Arbitrary Origin and Host values are
|
|
914
|
-
rejected. Each MCP session may read private context and memory only for the Agent selected by its
|
|
915
|
-
task entry. A checkpoint token is also bound to both the upstream's configured project and the MCP
|
|
916
|
-
session that created it, so neither another project process nor a peer session can reuse it.
|
|
199
|
+
<details>
|
|
200
|
+
<summary><b>Who decides what FULI keeps?</b></summary>
|
|
917
201
|
|
|
918
|
-
|
|
919
|
-
|
|
920
|
-
|
|
921
|
-
|
|
922
|
-
|
|
923
|
-
but only a real connected Claude conversation can prove that behavior.
|
|
202
|
+
You do: AI looks things up and does the work, and humans retain final authority. Each task ends
|
|
203
|
+
with one check that keeps only reusable knowledge. A preference an Agent guessed counts once you,
|
|
204
|
+
or <img src="docs/readme/fla-tonborg-inline.webp" alt="" width="20" align="absmiddle" /> Tonborg within the autonomy you set, confirm it. Older answers stay in the history with their
|
|
205
|
+
source and reason.
|
|
206
|
+
</details>
|
|
924
207
|
|
|
925
|
-
|
|
926
|
-
read-only `user-taste` Skill projection from the effective personal profile and prior task data.
|
|
927
|
-
The projection labels evidence status and scope, returns task-matched recommendations, and is
|
|
928
|
-
regenerated as preferences are added or revised. It never overwrites a user-authored taste Skill;
|
|
929
|
-
the durable graph remains the source of truth.
|
|
208
|
+
## Learn more
|
|
930
209
|
|
|
931
|
-
|
|
|
210
|
+
| Topic | Read |
|
|
932
211
|
| --- | --- |
|
|
933
|
-
|
|
|
934
|
-
|
|
|
935
|
-
|
|
|
936
|
-
|
|
|
937
|
-
|
|
|
938
|
-
|
|
|
939
|
-
|
|
|
940
|
-
|
|
|
941
|
-
|
|
|
942
|
-
|
|
943
|
-
|
|
944
|
-
|
|
945
|
-
| `preview_common_knowledge_promotion` | Preview a human-confirmed promotion |
|
|
946
|
-
| `apply_common_knowledge_promotion` | Apply the preview atomically with a one-time token |
|
|
947
|
-
| `resolve_deferred_preference_conflict` | Resolve an AI-deferred conflict only when a task needs it |
|
|
948
|
-
| `start_knowledge_review` | Start or resume an exact personal review scope |
|
|
949
|
-
| `list_knowledge_review_candidates` | Rank candidates by time, conflicts, weight, and cross-session repetition |
|
|
950
|
-
| `record_knowledge_review_progress` | Save keep, update, invalidate, skip, or defer outcomes |
|
|
951
|
-
| `finish_knowledge_review` | Pause, or complete and advance the next-review watermark |
|
|
952
|
-
|
|
953
|
-
## Privacy and safety boundaries
|
|
954
|
-
|
|
955
|
-
- Personal knowledge is written to the local Provider.
|
|
956
|
-
- Agents distill structured knowledge separately from private, adapter-visible conversation journals.
|
|
957
|
-
- Tokens, cookies, private keys, credentials, private contact details, temporary logs, and raw command
|
|
958
|
-
output must not enter the graph.
|
|
959
|
-
- Graphiti's remote-LLM path is disabled; embeddings are currently computed locally.
|
|
960
|
-
- Searches are bounded by personal space and project scope instead of mixing every project.
|
|
961
|
-
- External connectors never write to third-party sources; raw credentials are rejected from stored
|
|
962
|
-
binding configuration.
|
|
963
|
-
- Shared promotion requires auditable human confirmation. Agent usage evidence can promote an item
|
|
964
|
-
only as far as `agent_confirmed`.
|
|
965
|
-
- Fuli is not a real-time monitoring or Git service; current state must be read from the relevant
|
|
966
|
-
source system.
|
|
967
|
-
|
|
968
|
-
## Local services
|
|
969
|
-
|
|
970
|
-
| Service | Default loopback address | Availability |
|
|
971
|
-
| --- | --- | --- |
|
|
972
|
-
| Fuli management UI | `127.0.0.1:2727` | Always |
|
|
973
|
-
| Personal Provider | `127.0.0.1:8787` | Always |
|
|
974
|
-
| Personal Neo4j Browser / Bolt | `127.0.0.1:8060` / `127.0.0.1:7687` | Always |
|
|
975
|
-
| Development shared Provider | `127.0.0.1:8788` | `--with-dev-public` only |
|
|
976
|
-
| Development shared Neo4j Browser / Bolt | `127.0.0.1:7475` / `127.0.0.1:7688` | `--with-dev-public` only |
|
|
977
|
-
|
|
978
|
-
These seven ports can be changed in Settings. Provider and Neo4j bindings remain loopback-only even
|
|
979
|
-
when the management UI is started in LAN mode.
|
|
980
|
-
|
|
981
|
-
To access the UI from another device on the same Wi-Fi, run `fuli start --lan`. Starting LAN mode
|
|
982
|
-
from a loopback-only runtime safely restarts only the management UI and preserves Provider and
|
|
983
|
-
graph data. Disable LAN in Settings and run `fuli restart`, or run `fuli restart --no-lan`, to return
|
|
984
|
-
to loopback-only mode. Starting LAN mode again rotates the temporary access code.
|
|
985
|
-
|
|
986
|
-
If a port is occupied, Docker Compose is unavailable, or the container runtime is not running,
|
|
987
|
-
setup stops before modifying agent configuration and reports the cause.
|
|
988
|
-
|
|
989
|
-
## Acceptance and development
|
|
990
|
-
|
|
991
|
-
- [Alignment Benchmark](FULI_ALIGNMENT_BENCHMARK.md)
|
|
992
|
-
- [Chinese acceptance index](acceptance/README.md)
|
|
993
|
-
- [Knowledge retrieval and confirmation diagrams](acceptance/知识检索与确认流程图.md)
|
|
994
|
-
- [Read-only external-knowledge architecture](docs/external-knowledge-architecture.md)
|
|
995
|
-
|
|
996
|
-
```bash
|
|
997
|
-
npm install
|
|
998
|
-
npm test
|
|
999
|
-
npm run test:package
|
|
1000
|
-
npm run test:external-knowledge:live
|
|
1001
|
-
```
|
|
1002
|
-
|
|
1003
|
-
The external-knowledge live test temporarily downloads two public official documentation
|
|
1004
|
-
repositories, validates read-only synchronization and retrieval, then removes the complete
|
|
1005
|
-
temporary directory. It is network-dependent and is not part of the default test suite.
|
|
1006
|
-
|
|
1007
|
-
Provider validation:
|
|
1008
|
-
|
|
1009
|
-
```bash
|
|
1010
|
-
python3.12 -m venv .venv
|
|
1011
|
-
. .venv/bin/activate
|
|
1012
|
-
python -m pip install "./graph-provider[dev]"
|
|
1013
|
-
python -m compileall -q graph-provider/fuli_graph
|
|
1014
|
-
python -m pytest -q graph-provider/tests
|
|
1015
|
-
docker compose -f compose.graphiti.yml config --quiet
|
|
1016
|
-
```
|
|
1017
|
-
|
|
1018
|
-
`npm run test:package` builds the web UI, creates a real npm tarball, installs it into an isolated
|
|
1019
|
-
global prefix, and verifies `fuli` / `fl`, version output, help output, and the packaged UI. Tests,
|
|
1020
|
-
QA screenshots, and internal design documents are excluded from the npm package.
|
|
1021
|
-
|
|
1022
|
-
## Brand
|
|
1023
|
-
|
|
1024
|
-
<picture>
|
|
1025
|
-
<source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/JevonsCode/fuli/main/brand/svg/lockup/fuli-lockup-bilingual-horizontal-dark.svg" />
|
|
1026
|
-
<img src="https://raw.githubusercontent.com/JevonsCode/fuli/main/brand/svg/lockup/fuli-lockup-bilingual-horizontal.svg" alt="FULI 复利 lockup" width="280" />
|
|
1027
|
-
</picture>
|
|
1028
|
-
|
|
1029
|
-
The FULI mark, wordmarks and lockups come from the [FULI 复利 brand system on Figma](https://www.figma.com/community/file/1698659778015746085). See [`brand/README.md`](brand/README.md) for the brand files and how to change them.
|
|
1030
|
-
|
|
1031
|
-
## License
|
|
1032
|
-
|
|
1033
|
-
[Apache-2.0](LICENSE)
|
|
212
|
+
| Install details, CLI commands, ports | [Install and run](docs/install.md) |
|
|
213
|
+
| Principles and what is stored where | [How FULI works](docs/how-it-works.md) |
|
|
214
|
+
| Connecting each client | [Connect a client](docs/client-connections.md) |
|
|
215
|
+
| Skills, hooks and MCP tools | [Agent integrations](docs/agent-tools.md) |
|
|
216
|
+
| Notion, Lark and other knowledge bases | [External knowledge](docs/external-knowledge.md) |
|
|
217
|
+
| <img src="docs/readme/fla-tonborg-inline.webp" alt="" width="20" align="absmiddle" /> Tonborg and autonomy | [Judgment and autonomy](docs/judgment-and-autonomy.md) |
|
|
218
|
+
| The roundtable | [Roundtable](docs/agent-roundtable.md) |
|
|
219
|
+
| What is proven so far | [Capabilities and limits](docs/capabilities-and-limits.md) · [Measured report](https://有点.意思.com/fuli/report/en/) |
|
|
220
|
+
| Running the source | [Develop and verify](docs/development.md) · [Contributing](CONTRIBUTING.md) |
|
|
221
|
+
|
|
222
|
+
The FULI mark and wordmarks come from the [FULI 复利 brand system](https://www.figma.com/community/file/1698659778015746085); see [`brand/README.md`](brand/README.md).
|
|
223
|
+
Licensed under [Apache-2.0](LICENSE).
|