mirror of
https://github.com/tiddly-gittly/TidGi-Desktop.git
synced 2026-01-21 12:02:57 -08:00
6384fd8bd1
* fix: not removed
* Optimize tidgi.config.json writes for workspace updates
Update logic to write tidgi.config.json only for the modified workspace instead of all wiki workspaces on each update. This reduces redundant file operations and improves performance during workspace updates.
* Refactor workspace saving and UI update logic
Introduced a private saveWorkspacesToSettings method to centralize logic for saving workspaces and removing syncable fields from wiki workspaces. The set and setWorkspaces methods now support skipping UI updates for batch operations, improving performance. Fixed minor issues in legacy migration and error messages.
* Add 'Ask AI' context menu and wiki embed split view
Introduces an 'Ask AI' option to the wiki context menu, enabling users to send selected text to an agent chat in a split view with the wiki embedded. Implements new tab type WIKI_EMBED, updates tab and channel types, adds localization, manages BrowserView bounds for embedding, and ensures persistence and IPC wiring for the new workflow.
* Update wiki
* electron chrome mcp mode sometimes wont show browser view
Clarified troubleshooting steps in docs/MCP.md regarding browser view issues and updated the instructions. Reordered the 'start:dev:mcp' script in package.json for better organization.
* Add agent selection to 'Talk with AI' context menu
Replaces the 'Ask AI' context menu with 'Talk with AI' and adds a submenu for selecting different agent definitions. Updates translations for all supported languages, modifies the askAIWithSelection channel to support agentDefId, and refactors tab creation logic to support split view with agent selection. Improves robustness in view management by handling case-insensitive workspace IDs and custom bounds logic.
* Add e2e test and refactor 'Talk with AI' split view logic
Introduces a new Cucumber feature for 'Talk with AI' from wiki selection, adds a step definition to trigger the workflow via IPC, and refactors split view tab creation to reuse existing tabs when possible. Updates the agent browser service to support finding or creating the appropriate split view tab, and adjusts menu and view services for improved robustness and code clarity. Also adds test IDs to relevant components for more reliable UI testing.
* Update defaultWiki.feature
* Add config error handling and i18n for agent errors
Introduces a new feature test for configuration error handling, adds step definition to remove AI settings for testing, and updates error message rendering to support new error types. Internationalized error messages and button labels for configuration issues are added in both English and Chinese locales. The error message renderer now uses a data-testid for easier testing and recognizes additional error types as fixable in settings.
* Refactor feature files to use two-column selector tables
Updated all feature files to use a standardized two-column format for selector tables, with explicit 'element description' and 'selector' columns. Step definitions in ui.ts were refactored to support this format, improving readability and maintainability of test steps and error handling.
* Delete tiddlywiki
* test: allow parallel
* test: implement scenario isolation for E2E tests
- Isolate each test scenario in test-artifacts/{scenarioSlug}/ directory
- Use dynamic ports for mock OpenAI server to avoid port conflicts
- Log VIEW_LOADED event via did-finish-load in main process (more reliable)
- Search all .log files when waiting for log markers
- Increase timeout for log marker steps to 15 seconds
- Fix ts-node cache issues by clearing cache before tests
- Move application launch to individual scenarios (required for mock server setup)
All 45 E2E test scenarios now pass consistently.
* refactor: optimize agent.feature by moving common steps to Background
- Add MockOpenAIServer.addRules() method to append responses dynamically
- Add 'I have started the mock OpenAI server without rules' step for Background
- Add 'I add mock OpenAI responses:' step to inject responses per scenario
- Move application launch and navigation to Background (shared by all scenarios)
- Keep scenario-specific mock responses in individual scenarios
This improves test maintainability by reducing duplication while keeping
scenario-specific configuration flexible.
* lint
* Refactor scenario path helpers into shared module
Moved scenario-specific path helper functions from individual step definition files to a centralized 'features/supports/paths.ts' module. Updated imports in step definitions to use the shared helpers, improving code reuse and maintainability. Also enhanced test for ContextService to skip optional runtime keys.
* Refactor slug generation to use shared slugify helper
Introduced a new src/helpers/slugify.ts utility for consistent slug generation across the codebase. Updated appPaths.ts to use the shared slugify function, improving maintainability and ensuring identical behavior for test scenario slugs. Added documentation and clarified slugification rules in relevant files. Minor comments and clarifications were added to E2E and mock server code.
* Enforce strict timeout rules in E2E test steps
Added and clarified critical warnings for AI agents regarding timeout modifications in application, cleanup, and wiki step definitions. All timeouts are now strictly limited to 5s local and 10s CI, with explicit comments and environment-based values. Updated documentation and code comments to reinforce that timeouts indicate real bugs and should not be increased.
* Update features/stepDefinitions/application.ts
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Refactor E2E helpers, improve test reliability and cleanup
Centralizes data table parsing for UI step definitions, refactors mock OpenAI server setup, and improves workspace settings path handling for tests. Adjusts timeouts for window and app closing to better reflect real-world performance. Fixes type usage in workspace ID lookups and adds error handling for resize observer and cleanup in WikiEmbedTabContent. Enhances agent browser tab logic and view service cleanup to prevent memory leaks and catch workspace ID casing issues.
* Update agent.ts
* fix: resolve all E2E test timeout issues
* fix: improve CI test reliability with better timing and cleanup
- Use exponential-backoff library for agent creation retry logic
- Extend agent cancel delay to 1000ms for CI environments
- Fix git log refresh marker timing with queueMicrotask
- Improve cleanup timeout handling with force close strategy
All E2E tests passing locally including previously failing CI tests.
* Improve Git log E2E signal and add debug logging
Renames the test artifact in the CI workflow for clarity. Moves the E2E test timing log in useGitLogData to after entries are rendered, using a more reliable signal. Adds a debug log to notifyGitStateChange for better traceability.
* test-artifacts-ci
* Optimize test artifact handling and Git log logging
Update CI workflow to clean up large cache folders in test artifacts and only upload logs, settings, and screenshots to reduce artifact size. Refactor useGitLogData to log immediately after state updates for improved E2E test reliability, removing unnecessary setTimeout.
* Update useGitLogData.ts
* Improve Git log E2E test logging and .gitignore
Added 'test-artifacts-ci.zip' to .gitignore. Moved the '[test-id-git-log-refreshed]' log to immediately after data load for more reliable E2E test detection, and removed redundant logging from the render effect in useGitLogData.ts.
* Update useGitLogData.ts
* Update useGitLogData.ts
* Fix git log refresh marker not appearing in CI
- Move git-log-refreshed marker before RAF to ensure it's recorded
- RAF callbacks may not execute reliably in headless CI environments
- Add debug logging to track loadGitLog execution
- Add try-catch around log call to catch any errors
- Keep git-log-data-rendered in useEffect for UI tracking
* Update useGitLogData.ts
* Update useGitLogData.ts
* Add comprehensive logging to diagnose git-log-refreshed issue
- Log before RAF and inside RAF to pinpoint exact failure location
- Add try-catch to capture any errors
- Two log markers: before-raf and in-raf
- This will definitively show where the logging fails in CI
* Fix race condition: prevent concurrent loadGitLog calls
Root cause: commit triggers 2 refreshes (gitStateChange$ + handleCommitSuccess)
- First loadGitLog (refreshTrigger=1) succeeds
- Second loadGitLog (refreshTrigger=2) starts but never completes
- Add loadGitLogInProgress guard to prevent concurrent execution
- Log when loadGitLog is skipped due to in-progress call
This ensures git-log-refreshed is always logged after commit.
* Remove redundant triggerRefresh calls causing race condition
- handleCommitSuccess/Revert/Undo no longer call triggerRefresh
- gitStateChange\$ observable already triggers refresh for these operations
- Redundant calls caused 2 concurrent loadGitLog, causing CI test failures
- Local tests passed because both completed; CI failed because 2nd never completed
This ensures only 1 loadGitLog runs per git operation.
* Remove unused triggerRefresh parameter from useCommitSelection
- triggerRefresh no longer used in handlers
- Remove from interface and call site
- Clean up lint errors
* Remove triggerRefresh completely - no longer needed
- Observable subscription handles all git state changes
- Remove function definition and exports
- Fix all lint errors
Root cause resolved: commit triggered double refresh causing race condition.
Now only single refresh via observable.
* Remove fixed time waits from gitLog.feature and fix race condition
- Remove all fixed time wait steps from gitLog.feature (14 instances)
- Remove redundant triggerRefresh calls in handleCommitSuccess/Revert/Undo
- Add loadGitLogInProgress guard to prevent concurrent loadGitLog
- Root cause: commit triggered 2 refreshes causing race condition
- Only gitStateChange\$ observable now triggers refresh
- All 4 gitLog tests pass locally
* Fix clear timing: clear log BEFORE commit, not after
Root cause: test cleared git-log-refreshed AFTER commit completed
- But commit already triggered refresh and logged git-log-refreshed
- Clear deleted it, then test waited for new log that would never come
- Solution: clear BEFORE clicking commit button
- This way commit's git-log-refreshed is the first one after clear
Test now passes locally.
* Update cleanup.ts
* Initial commit when init a new git.
* Refactor feature steps for multi-element and log marker tables
Updated multiple feature files and step definitions to support table-driven steps for clicking and asserting multiple elements, and for waiting for multiple log markers in sequence. This reduces redundant waits, improves test reliability, and streamlines Gherkin syntax for multi-element actions and assertions. Also removed unnecessary manual wait steps where content or element checks now handle waiting automatically.
* Minor code cleanup and formatting improvements
Reordered imports in browserView.ts, fixed whitespace in cleanup.ts and useGitLogData.ts, and improved line formatting in GitLog/index.tsx for better readability and consistency.
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: tidgi <tiddlygit@gmail.com>
128 lines
10 KiB
Gherkin
128 lines
10 KiB
Gherkin
Feature: Agent Workflow - Tool Usage and Multi-Round Conversation
|
||
As a user
|
||
I want to use an intelligent agent to search wiki content
|
||
So that I can get AI-powered explanations of wiki entries
|
||
|
||
Background:
|
||
Given I add test ai settings
|
||
# Start mock OpenAI server with no rules - rules will be added per scenario
|
||
And I have started the mock OpenAI server without rules
|
||
Then I launch the TidGi application
|
||
And I wait for the page to load completely
|
||
And I should see a "page body" element with selector "body"
|
||
# Navigate to agent workspace for all scenarios
|
||
And I click on "agent workspace button and new tab button" elements with selectors:
|
||
| element description | selector |
|
||
| agent workspace | [data-testid='workspace-agent'] |
|
||
| new tab button | [data-tab-id='new-tab-button'] |
|
||
|
||
@agent @mockOpenAI
|
||
Scenario: Wiki-search tool usage
|
||
# Add scenario-specific responses to the mock server
|
||
Given I add mock OpenAI responses:
|
||
| response | stream |
|
||
| <tool_use name="wiki-search">{"workspaceName":"-VPTqPdNOEZHGO5vkwllY","filter":"[title[Index]]"}</tool_use> | false |
|
||
| 在 TiddlyWiki 中，Index 条目提供了编辑卡片的方法说明，点击右上角的编辑按钮可以开始对当前卡片进行编辑。此外，它还引导您访问中文教程页面和官方英文站点以获取更多信息。 | false |
|
||
# Proceed with agent workflow in main window
|
||
# Step 1: Click new tab button
|
||
When I click on a "new tab button" element with selector "[data-tab-id='new-tab-button']"
|
||
And I should see a "search interface" element with selector ".aa-Autocomplete"
|
||
# Step 2: Click search box and wait for autocomplete
|
||
When I click on a "search input box" element with selector ".aa-Input"
|
||
And I should see an "autocomplete panel" element with selector ".aa-Panel"
|
||
# Step 3: Select agent from autocomplete (not new tab)
|
||
When I click on an "agent suggestion" element with selector '[data-autocomplete-source-id="agentsSource"] .aa-ItemWrapper'
|
||
And I should see a "message input box" element with selector "[data-testid='agent-message-input']"
|
||
# Step 4: Send message to agent - using generic steps combination
|
||
When I click on a "message input textarea" element with selector "[data-testid='agent-message-input']"
|
||
When I type "搜索 wiki 中的 index 条目并解释" in "chat input" element with selector "[data-testid='agent-message-input']"
|
||
And I press "Enter" key
|
||
Then I should see 4 messages in chat history
|
||
# Verify the last message contains the AI explanation about Index
|
||
And I should see "explanation in last message and explanation about edit" elements with selectors:
|
||
| element description | selector |
|
||
| explanation in last message | [data-testid='message-bubble']:last-child:has-text('Index') |
|
||
| explanation about edit | [data-testid='message-bubble']:last-child:has-text('编辑') |
|
||
|
||
@agent @mockOpenAI
|
||
Scenario: Wiki operation
|
||
# Add scenario-specific responses to the mock server
|
||
Given I add mock OpenAI responses:
|
||
| response | stream |
|
||
| 先测试失败情况<tool_use name="wiki-operation">{"workspaceName":"test-expected-to-fail","operation":"wiki-add-tiddler","title":"testNote","text":"test"}</tool_use> | false |
|
||
| 然后测试成功情况<tool_use name="wiki-operation">{"workspaceName":"wiki","operation":"wiki-add-tiddler","title":"test","text":"这是测试内容"}</tool_use>使用启动时自动创建的 wiki 工作区 | false |
|
||
| 已成功在工作区 wiki 中创建条目 "test"。 | false |
|
||
# Step 1: Start a fresh tab and run the two-round wiki operation flow
|
||
When I click on a "new tab button" element with selector "[data-tab-id='new-tab-button']"
|
||
And I should see a "search interface" element with selector ".aa-Autocomplete"
|
||
# Step 2: Click search box and wait for autocomplete
|
||
When I click on a "search input box" element with selector ".aa-Input"
|
||
And I should see an "autocomplete panel" element with selector ".aa-Panel"
|
||
# Step 3: Select agent from autocomplete (not new tab)
|
||
When I click on an "agent suggestion" element with selector '[data-autocomplete-source-id="agentsSource"] .aa-ItemWrapper'
|
||
And I should see a "message input box" element with selector "[data-testid='agent-message-input']"
|
||
# First round: try create note using test-expected-to-fail workspace (expected to fail)
|
||
When I click on a "message input textarea" element with selector "[data-testid='agent-message-input']"
|
||
When I type "在 wiki 里创建一个新笔记，内容为 test" in "chat input" element with selector "[data-testid='agent-message-input']"
|
||
And I press "Enter" key
|
||
Then I should see 6 messages in chat history
|
||
# Verify error and success messages
|
||
And I should see "workspace not exist error and success in last message and wiki workspace in last message" elements with selectors:
|
||
| element description | selector |
|
||
| workspace not exist error | [data-testid='message-bubble']:has-text('test-expected-to-fail'):has-text('不存在') |
|
||
| success in last message | [data-testid='message-bubble']:last-child:has-text('已成功') |
|
||
| wiki workspace in last message | [data-testid='message-bubble']:last-child:has-text('wiki') |
|
||
|
||
@agent
|
||
Scenario: Create default agent from New Tab quick access
|
||
When I click on "new tab button and create default agent button" elements with selectors:
|
||
| element description | selector |
|
||
| new tab button | [data-tab-id='new-tab-button'] |
|
||
| create default agent button | [data-testid='create-default-agent-button'] |
|
||
And I should see a "message input box" element with selector "[data-testid='agent-message-input']"
|
||
|
||
@agent
|
||
Scenario: Close all tabs then create default agent from fallback page
|
||
# Ensure starting from black/fallback page with no open tabs
|
||
Given I click on a "new tab button" element with selector "[data-tab-id='new-tab-button']"
|
||
When I click all "tab" elements matching selector "[data-testid='tab']"
|
||
When I click all "close tab button" elements matching selector "[data-testid='tab-close-button']"
|
||
# When there is no active tab, this is "fallback new tab", it has same thing as new tab.
|
||
And I should see "new tab button and Create Default Agent" elements with selectors:
|
||
| element description | selector |
|
||
| new tab button | [data-tab-id='new-tab-button'] |
|
||
| Create Default Agent | [data-testid='create-default-agent-button'] |
|
||
When I click on a "new tab button" element with selector "[data-tab-id='new-tab-button']"
|
||
And I should see a "Create Default Agent" element with selector "[data-testid='create-default-agent-button']"
|
||
When I click on a "create default agent button" element with selector "[data-testid='create-default-agent-button']"
|
||
And I should see a "message input box" element with selector "[data-testid='agent-message-input']"
|
||
Then I click all "close tab button" elements matching selector "[data-testid='tab-close-button']"
|
||
|
||
@agent @mockOpenAI
|
||
Scenario: Streamed assistant response can be cancelled mid-stream and send button returns
|
||
# Add scenario-specific responses to the mock server
|
||
Given I add mock OpenAI responses:
|
||
| response | stream |
|
||
| partial_chunk_1<stream_split>partial_chunk_2<stream_split>partial_chunk_3<stream_split>partial_chunk_4 | true |
|
||
And I click on "new tab button and create default agent button" elements with selectors:
|
||
| element description | selector |
|
||
| new tab button | [data-tab-id='new-tab-button'] |
|
||
| create default agent button | [data-testid='create-default-agent-button'] |
|
||
And I should see a "message input box" element with selector "[data-testid='agent-message-input']"
|
||
When I click on a "message input textarea" element with selector "[data-testid='agent-message-input']"
|
||
When I type "Start long streaming" in "chat input" element with selector "[data-testid='agent-message-input']"
|
||
And I press "Enter" key
|
||
# Wait for streaming container to appear and contain the first chunk
|
||
Then I should see "assistant streaming container and partial assistant text and cancel icon" elements with selectors:
|
||
| element description | selector |
|
||
| assistant streaming container | [data-testid='assistant-streaming-text'] |
|
||
| partial assistant text | *:has-text('partial_chunk_1') |
|
||
| cancel icon | [data-testid='cancel-icon'] |
|
||
# Click cancel button mid-stream
|
||
When I click on a "cancel button" element with selector "[data-testid='agent-send-button']"
|
||
# Verify send button returned and stream stopped (no further chunks)
|
||
Then I should see "send icon and send button" elements with selectors:
|
||
| element description | selector |
|
||
| send icon | [data-testid='send-icon'] |
|
||
| send button | [data-testid='agent-send-button']|
|
||
And I should not see a "partial chunk 4 text" element with selector "text='partial_chunk_4'"
|