Add explicit instructions across all agent prompts explaining that binary
files (PDFs, images, documents) return content as attachments in the
message following the tool result, not as text in the result itself.
Update tool response messages to clearly state the attachment delivery
mechanism and prevent agents from re-reading the same file expecting
different output.
Improve build safety by neutralizing dynamic eval constructs and
hardening the officeparser plugin against bundle shape changes.
Remove deprecated model versions (Claude 4/4.5/4.6, Gemini 2.5, GPT 5.4 Pro, Mistral latest aliases) and replace with current model identifiers. Add DEFAULT_QUICK_MODEL_BY_PROVIDER configuration, update default models to use Sonnet 4.6 standard and Opus 4.8 for planning, and adjust Mistral agent to use Small model.
- Implement "Suggest tags" action that generates and merges tags using AI
- Implement "Generate frontmatter" action that creates metadata fields
- Add updateFrontmatter method to VaultService and FileSystemService
- Create FrontmatterHelpers for YAML parsing and field merging
- Update prompts to reuse existing vault tags for consistency
- Remove deprecated mergeTagsIntoFrontmatter function
- Add yaml package dependency for frontmatter parsing
Move replaceCopy function from Copy enum to Helpers module to improve
separation of concerns. Add splitFrontmatter and mergeTagsIntoFrontmatter
utilities for YAML frontmatter manipulation. Remove unused Beautify prompt
and QuickActionsService. Add timeout support to Semaphore.wait(). Remove
unused Event.QuickActionsSettingsChanged. Update all imports across AI
classes, prompts, services, and tests.
Fix reactive issue for web and chat mode display.
Rewrites StreamingMarkdownService to use Obsidian's built-in MarkdownRenderer
instead of the unified/remark/rehype/KaTeX stack, eliminating ~700 lines of
custom markdown processing, CSS, and streaming state management. Internal link
handling now relies on Obsidian's native .internal-link class and data-href
attributes. The streaming split-point algorithm (freeze completed blocks, re-render
the live tail) is preserved.
Replace activeDocument/activeWindow globals with Obsidian API equivalents (createFragment, createDiv, createSpan, window), update officeparser to use async API, switch diff2html import path and disable highlight option, prefix unused parameters with underscore, and update dependencies including @anthropic-ai/sdk, @google/genai, officeparser, and various dev dependencies
- Change ChatMode enum values from numeric (0,1,2) to string literals ("read_only", "edit", "planning")
- Move chat mode state management from ChatWindow to SettingsService with persistence
- Add chat mode awareness to system prompts and tool validation
- Replace Map with WeakMap for settings subscribers to prevent memory leaks
- Add type-safe event handlers to EventService
- Update esbuild and TypeScript targets from ES2018/ES2020 to ES2022
- Add requiresEditModeEnabled() validation to prevent tool hallucination in read-only mode
- Update all test fixtures to include chatMode in mock settings
Mock the new subscribeToSettingsChanged method in all AI provider test files to prevent test failures. Update SettingsService tests to use updateSettings method instead of direct settings modification and saveSettings calls, aligning with the new reactive settings API that notifies subscribers of changes.
Add settings to enable/disable quick actions in both the editor context menu and toolbar. Implement dynamic registration system that updates when settings change. Include event-driven architecture for settings changes and improve toolbar injection logic. Update dependencies to latest versions.
Update test mocks to use renamed methods: readFile → readFilePath, writeFile → writeToFilePath, patchFile → patchFileAtPath. Add activeWindow and activeDocument globals to test setup for Obsidian compatibility.
Introduce QuickActionsService and QuickAgent for lightweight, single-shot AI
operations. Add a dedicated quickActionModel setting alongside a new top-level
provider setting, with validation ensuring all models match the selected provider
and provider-specific defaults. Refactor FileSystemService to offer both TFile
and path-based overloads (readFile/readFilePath, writeToFile/writeToFilePath,
patchFile/patchFileAtPath). Replace window/document globals with activeWindow/
activeDocument throughout InputService and VaultkeeperAISettingTab for Obsidian
mobile compatibility.
Update linting packages to latest.
- Calculate height for consecutive assistant messages instead of just last message
- Track message role in ChatArea for proper height calculation
- Update MultiAgentIntegration test to use ChatMode enum
- Update officeparser browser bundle path and shim for v6
Introduce a new setting to control AI access to the web viewer tool and refactor prompt building to inject capability directives (memories, web search, web viewer) into all agent contexts. Update tool definitions to conditionally include web viewer and memory tools based on settings. Consolidate memories injection and active directives into a unified prompt builder pattern.
Implement web search toggle functionality across all AI providers (Claude, Gemini, Mistral, OpenAI) with conditional tool inclusion based on settings. Add globe button to chat input for toggling web search on/off. Change formatBinaryFiles visibility from public to protected in BaseAIClass and IAIClass interface. Extract tool configuration into private getTools() methods for each provider. Update tests to verify web search toggle behavior and formatBinaryFiles access level changes.
- Add MoveVaultFolder tool with supporting schema and enum
- Rename schemas from *Schema to *ArgsSchema for consistency
- Fix VaultService.move() to handle folder destinations properly
- Update AIToolResponse to use AIToolResponsePayload wrapper
- Update error message for non-existent move sources
- Add MoveVaultFolder import and tool registration
Refactor all test files to wrap tool response data in AIToolResponsePayload instances instead of passing raw objects directly to AIToolResponse constructor. Update assertions to access response data via payload.response property.
Add new AI tool to create vault folders with proper permission checks and error handling. Refactor createDirectories to be public and handle both file and directory paths. Update WriteVaultFile description formatting and fix method naming consistency.
- Update error message to show actual max lines and words per line
- Return error message from updateMemories instead of success boolean
- Add SettingsService mocks to all agent tests to prevent initialization errors
- Update test expectations for new default search limits (30 results, 100 snippet size)
- Remove unused `path` property from Conversation class and its test
- Initialize `items` and `conversations` arrays in ConversationHistoryModal
- Use `Exception.messageFrom()` for safer error serialization in StreamingMarkdownService
- Fix redundant type cast in obsidian mock's Events.on()
- Bump @anthropic-ai/sdk, @google/genai, openai, esbuild, svelte, typescript, vitest, and other minor deps
Replace hardcoded "provided below" strings with "included below" and migrate to centralized Copy enum system for consistent messaging across Claude, Gemini, Mistral, and OpenAI implementations.
Add support for DOCX, PPTX, XLSX, ODT, ODP, and ODS document formats by converting them to plain text. Standardize attachment introduction messages across all AI providers from "Binary data for X follows" to "The contents of the file 'X' are provided below."
Implement complete Mistral AI support including message formatting, tool calls with ID validation (9-char alphanumeric), streaming, file attachments, and conversation naming service. Add extensive cross-provider integration tests covering Mistral interactions with Claude, OpenAI, and Gemini, including round-trip conversions, ID validation, and mixed conversation handling. Update all test fixtures to include Mistral API keys.
Enable PatchVaultFile to apply multiple patches in a single operation by converting oldContent and newContent from strings to arrays. Add whitespace-flexible regex matching as fallback when exact string matching fails, improving robustness for code with varying indentation. Update schemas, tool definitions, and all tests to reflect the new array-based API.
Replace the single Replan tool with three targeted alternatives:
ReviseStep, RevisePlan, and SkipStep. This gives the orchestration
agent more precise control over plan recovery — revising only the
current step, replacing all remaining steps, or skipping a step
entirely — rather than triggering a full replan each time.
Also give the orchestration agent read access to vault files so it
can resolve execution failures without routing back to the planning
agent, and update the execution agent prompt to stop it from
attempting self-recovery (gap resolution is now the orchestrator's
responsibility).
Add getMimeType method mock to all Attachment test objects across AI service tests (Claude, Gemini, OpenAI) and mock StringTools.resizeB64Image in BaseAIFileService tests to prevent browser environment issues.
- Rename toolDefinitions to aiFunctionDefinitions across all AI classes
- Add aiFunctionUsageMode property to control function calling behavior
- Implement provider-specific tool_choice/tool_config based on usage mode
- Remove AIController and create BaseAgent base class
- Update ExecutionPrompt with scope of execution guidelines
- Simplify ChatArea layout update logic by removing debounce
- Add naming service completion wait in ChatService
- Replace console.warn with Exception.warn in InputService
- Delete unused AIController.ts file
- Update all tests to use new aiFunctionDefinitions property
- Rename AIControllerService to MainAgent throughout test files
- Move AIFunctionService to AIServices directory
- Update service registration from AIControllerService to MainAgent
- Add agentType parameter to mock AI configuration
- Add resolveAIProvider method to mock service
- Update test comments to reflect orchestration agent workflow
- Remove planning/execution agent-specific tests in favor of integration tests
- Consolidate agent loop tests into main test suite
- Add context_for_next_step parameter to CompleteStep for passing execution history
- Add context parameter to CompleteTask for preserving task completion state
- Update OrchestrationResult to handle context propagation between steps
- Add debug color differentiation for agent types (Main, Execution, Orchestration, Planning)
- Reorganize SearchTypes from Helpers to Types directory
- Add justification requirement for execution deviations
- Support reasonable deviations in orchestration plan validation
- Refactor dependency service with TryResolve utility
- Add whitespace cleanup to Semaphore class
Restructure the AI workflow from a single-agent model to a specialized
multi-agent system with distinct roles:
- Add AgentType enum (Main, Orchestration, Planning, Execution) to define
agent specializations
- Replace AIControllerService and AIFunctionService with modular agent
classes in Services/AIServices/:
- MainAgent: Handles user interaction and delegates to orchestration
- OrchestrationAgent: Coordinates plan execution and step-by-step workflow
- PlanningAgent: Creates and revises execution plans
- ExecutionAgent: Executes individual plan steps
- AIController: Base class providing common agent loop functionality
- Create specialized prompts (OrchestrationPrompt, ExecutionPrompt) for
agent-specific behavior
- Add OrchestrationResult type to communicate workflow control decisions
(continue, abort, replan)
- Introduce agent-specific function scoping:
- ExecuteWorkflow for main agent
- CompleteTask for execution agent
- CompleteStep/Replan/CancelPlan for orchestration agent
- SubmitPlan/AskUserQuestionPlanning for planning agent
- Update BaseAIClass to use agentType property instead of isPlanningAgent
boolean for model selection
- Simplify ExecutionPlan and ExecutionStep types by removing execution
state tracking (moved to agent coordination)
- Remove PlanningEnabledAppendix and ExecutionStatus enum (superseded by
agent architecture)
- Add comprehensive integration tests for agent workflows
This architecture provides better separation of concerns, clearer agent
responsibilities, and more robust plan execution with explicit orchestration
control flow.
Test agent loop function calls, planning mode, execution agent, replan workflow, error handling, save callbacks, and shouldContinue flag behavior. Update mocks to use AIFunctionResponse class and rename onPlanComplete to onPlanReset.
Add distinct AskUserQuestionPlanning and AskUserQuestionExecution functions to enable user consultation during both planning and execution phases. Update system prompt to clarify when to seek user input vs. replan. Fix plan area rendering timing and improve execution flow handling.
- Add currentProvider getter to AI classes for provider access
- Enhance planning prompt to discourage over-engineering and scope creep
- Return Unknown enum instead of throwing for invalid AI functions
- Add completion reminders to execution steps
- Fix ChatPlanArea height calculation timing
- Update tests to handle Unknown function gracefully
