This repository was archived by the owner on Mar 7, 2026. It is now read-only.
feat: onboarding redesign, memory conflict resolution, identity fix#6
Merged
feat: onboarding redesign, memory conflict resolution, identity fix#6
Conversation
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Wraps ModelFetcher.fetchModels() to return ValidationResult, classifying probe outcomes into Success/Failure/Offline/Idle for onboarding and settings UI. Includes 9 unit tests covering all classification branches. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Validates Telegram, Discord, Slack, and Matrix tokens by probing their respective APIs. Other channel types return a deferred success result. Classification logic is exposed as internal functions for unit testing without network access. All 12 tests pass. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Centralises deep-link URIs and Intent construction for onboarding and settings flows. Includes DeepLinkTarget data class, pre-defined targets for Telegram, Discord, and Slack, provider console lookup for 7 AI providers, and a launch function with fallback URI handling. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Define a data-driven specification for interactive channel setup with per-platform step sequences, instructions, deep links, field grouping, and validator types. Includes dedicated specs for Telegram, Discord, Slack, and Matrix, plus generic two-step specs for all other channel types. Fields reference existing ChannelType.fields instances to prevent drift. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…omposables Add three reusable setup-flow composables in ui/component/setup/: - ValidationIndicator: renders ValidationResult states with icons/spinners - InstructionsList: renders InstructionItem list with numbered steps, warnings, hints - DeepLinkButton: OutlinedButton that opens external apps via ExternalAppLauncher Also fix pre-existing detekt issues: suppress TooGenericExceptionCaught in ChannelValidator catch blocks, bump TooManyFunctions threshold for objects, and configure UnusedPrivateMember to ignore @Preview-annotated functions. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Extract shared provider setup UI (credential form, validation button, deep-link to console, model suggestions) into ProviderSetupFlow in the setup component package. ProviderStep becomes a thin wrapper that adds onboarding-specific title/description and live model fetching logic. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ChannelSetupFlow renders per-channel sub-step wizards with instructions, deep links, dynamic field rendering, and validation support. Fields are mapped from FieldInputType to appropriate keyboard types and widgets including secret fields with reveal toggles and boolean switches. ChannelSelectionGrid provides a two-column multi-select grid for choosing which channels to configure, with visual indicators for selected and already-configured channels. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…ntityConfigFlow, ConfigSummaryCard Add five reusable setup composables for the onboarding redesign: - TunnelConfigFlow: tunnel type selector with conditional token/URL fields - AutonomyPicker: three-level autonomy selector with warning styling - MemoryConfigFlow: backend, auto-save, embedding provider, retention - IdentityConfigFlow: agent name, user name, timezone, style, format - ConfigSummaryCard: read-only overview with status icons per dimension All field names align with ConfigTomlBuilder/GlobalTomlConfig. Each composable includes KDoc, accessibility semantics, and preview variants. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace the 5-step OnboardingViewModel with a 9-step OnboardingCoordinator that supports multiple channel sub-flows, tunnel configuration, security/ autonomy settings, memory backend selection, and identity customisation. New files: - OnboardingStepStates.kt: Data classes for each step's state - OnboardingCoordinator.kt: AndroidViewModel with 9-step navigation, provider validation, channel sub-flow management, debounced model fetching, and enhanced completion logic preserving all existing OnboardingViewModel persistence - OnboardingCoordinatorTest.kt: 24 unit tests covering step navigation bounds, state defaults, copy semantics, and auth failure detection All existing completion logic is preserved: auth probing, API key saving, agent create/update, channel saving (now multi-channel), identity JSON, workspace scaffolding, provider/model defaults, and PIN/lock persistence. New persistence added for tunnel, autonomy, and memory configuration. Also applies spotless formatting corrections to setup composables. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Addresses spec review findings: - Expose welcomeState + acknowledgeWelcome() from coordinator - Inline isDefinitiveAuthFailure to remove coupling to OnboardingViewModel Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace OnboardingViewModel with OnboardingCoordinator in OnboardingScreen, expanding the wizard from 5 to 9 steps: permissions, welcome, provider, channels, tunnel, security, memory, identity, and activation. - Create WelcomeStep composable with hero icon and setup preview card - Create SecurityStep composable wrapping AutonomyPicker - Wire ChannelStepCollector with ChannelSelectionGrid and ChannelSetupFlow for per-channel interactive sub-flows - Add TunnelStepCollector, SecurityStepCollector, MemoryStepCollector, IdentityStepCollector, and ActivationStepCollector - Update ActivationStep to render ConfigSummaryCard with aggregated config - Update OnboardingContent to remove onActivate parameter since the activation step renders its own button - Update instrumented tests and FakeData for 9-step state Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace the channel type picker dialog with ChannelSelectionGrid for a richer visual selection experience. Add ChannelSetupFlow wizard for new channel creation in ChannelDetailScreen while preserving the flat form for editing existing channels. Add inline validation with ValidationIndicator to each API key card in ApiKeysScreen via ProviderValidator. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Mark the legacy 5-step OnboardingViewModel as @deprecated with a pointer to OnboardingCoordinator. Update KDoc references in AgentConfigStep and all test files to point to the new coordinator. All unit tests, spotless, detekt, and assembleDebug pass cleanly. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add three methods to DaemonServiceBridge: - detectMemoryConflict(): scans workspace for stale backend artifacts - cleanupStaleMemory(): deletes stale files (sqlite, markdown, or both) - checkMemoryHealth(): write/read probe to verify backend storage Includes 13 unit tests covering all backends, state/ subdirectory, and edge cases. Reviewed for spec compliance and code quality. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add memoryConflict and memoryHealthWarning StateFlows, plus awaitConflictResolution() suspend method and setMemoryHealthWarning() for the service to use during startup orchestration. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Insert conflict detection between config build and attemptStart() in handleStartFromSettings(). Suspends on awaitConflictResolution() when stale files are found. Post-start health check via runMemoryHealthCheck(). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…iewModel Add memoryConflict, memoryHealthWarning StateFlows and resolution methods so the dashboard UI can observe and respond to conflicts. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Show blocking AlertDialog when stale memory backend files are detected at daemon startup. Add dismissible error banner for failed health checks. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Console messages were routed through the stateless gateway webhook (run_gateway_chat_simple) which had no tools, no memory recall, and used native_tools=false causing hallucinated <tool_call> tags. Now routes through zeroclaw::agent::process_message directly, giving Console the same treatment as channels: memory recall, tool access, proper identity injection, and correct native_tools detection. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
send_messagethrough the full agent loop (process_message) instead of the stateless gateway webhook, giving Console messages memory recall, tool access, and proper workspace identity injectionKey changes
send_messagenow callszeroclaw::agent::process_messagedirectly — full agent loop with memory, tools, and identityTest plan
🤖 Generated with Claude Code