Add Task Recall documentation to both wizard prompts (wizard-system.md
and wizard-inline-system.md), enabling AI to access session history for
context about previously completed work. Extended documented fields to
include fullResponse, elapsedTimeMs, and contextUsage for detailed
context retrieval.
Updated wizardPrompts.ts and inlineWizardConversation.ts to pass
historyFilePath through template context. Modified useInlineWizard.ts
to fetch history file path during wizard initialization and store in
state for use across conversation session creation points.
Introduces an "About Me" field in Settings → General that allows users
to describe their background, preferences, and communication style.
This profile is exposed as {{CONDUCTOR_PROFILE}} template variable and
included in AI chat, group chat, and wizard prompts so agents can
tailor their responses to the user.
- Wizard now auto-continues when AI implies “let me research…” dead-ends 🧠
- Added robust deferred-response phrase detection to keep conversations flowing 🔍
- Introduced “Your turn” indicator when assistant awaits user input 💡
- Updated wizard system prompt to forbid deferred/async promises outright 🚫
- Prevented auto-continue infinite loops with a per-exchange trigger guard 🛡️
- Improved auto-continue UX with delayed send after UI updates ⏱️
- Added full test suite for deferred-response detection patterns ✅
- Strengthened NewInstanceModal tests with remote directory stat validation 📁
- NewInstanceModal tests now wait for debounced remote path validation ⏳
- Expanded icon mocking to include Loader2 for more complete test coverage 🎭
- SSH command builder is now async and resolves real ssh path 🚀
- Packaged Electron SSH launches now use expanded PATH environment 🧭
- Remote git and shell execution now await SSH command construction 🧰
- IPC process handler updated to await SSH wrapper command generation ⚙️
- Integration SSH test runner updated for async SSH command building 🌐
- Added SSH CLI detection with caching, sync getter, and reset hook 🔎
- Test suite now mocks SSH path resolution for deterministic runs 🧪
- Process IPC tests include robust async ssh-command-builder mock 🎭
- SSH builder logging now reports resolved binary path and full command 📝
- Prompts now enforce structured Markdown artifacts with front matter links 📚
- Wizard now saves generated docs under `Auto Run Docs/Initiation/` for clarity 📁
- Onboarding copy refresh: “Auto Run Playbook” terminology replaces “document” consistently 📝
- Claude parser now extracts and streams extended “thinking” blocks reliably 🧠
- Added full test coverage for thinking, redaction, and tool-use parsing paths ✅
- Wizard modal gains a real focus trap for rock-solid keyboard navigation ⌨️
- Wizard buttons now show accessible focus rings matching theme colors 🎯
- Phase Review: Escape closes document dropdown before modal closes 🛡️
- Phase Review: cycle multiple generated documents with ⌘/Ctrl+Shift+[ ] shortcuts 🔁
- Directory step: Enter on Browse triggers browse, not accidental “Continue” 🗂️
- Auto Run prompts upgraded with token-efficient task grouping rules and examples ⚡
- Added Claude “prompt too long” detection as recoverable token exhaustion 🧩
- Rebranded wizard output from “action plans” to “Playbooks” everywhere 📘
- AutoRun Edit/Preview controls are now icon-only with clearer titles 🎛️
- Introduced Playbook Exchange button styling and richer tooltip copy 🧺
- Wizard now routes chat input directly into its own conversation stream 💬
- Wizard completion now appends conversation logs and a next-steps summary 🧾
- Wizard captures Claude `session_id` and switches tab for continuity 🔗
- Wizard mode becomes tab-scoped, preventing cross-tab activation leakage 🧭
- Confidence gauge now withholds green until readiness threshold (80+) 🎨
- Document generation shows centered Austin Facts during initial loading 🧠
- Claude agents now run read-only tools for safer Playbook generation 🔒
- Wizard pill shows “Thinking…” spinner and pauses pulse while waiting ⏳
- Introduced explicit read/write access rules: Auto Run writes only! 🔒
- Updated all wizard prompt templates with clearer safety constraints 🧭
- Refined tests to validate new “WRITE Limited / READ Unrestricted” wording 🧪
- Synced inline wizard context into `session.wizardState` for stable UI 🧷
- Scoped inline wizard to the tab it started on for isolation 🧩
- Auto-sent an initial wizard greeting to kickstart discovery instantly 🚀
- Added “Exit wizard” support (Escape/pill) via new `onExitWizard` callback 🚪
- Fixed wizard loading indicator to use `wizardState.isWaiting` accurately ⏳
- Restricted wizard conversation tools to read-only operations for safety 📚
- Deduplicated parsing by reusing shared structured-output + confidence utilities 🧰
- Removed PROJECT_FILES_CONTEXT from wizard system prompt template 🗑️
- Added instruction to examine working directory before responding 🔍
- Simplified discovery approach to look for file patterns directly 📁
- Removed ProjectFileInfo interface and related functionality 🧹
- Cleaned up projectFiles parameter from configuration interfaces ✨
- Removed buildProjectFilesContext function and file analysis logic 🎯
- Updated system prompt to rely on direct file examination 👀
- Streamlined template variable replacement in prompt generation 🚀
- Maintained existing docs functionality for session continuity 📝
- Improved clarity of file-based project type detection approach 💡
Move all built-in prompts from inline code to separate .md files in src/prompts/
for easier editing without code changes. Prompts use {{TEMPLATE_VARIABLES}} that
are substituted at runtime using the central substituteTemplateVariables function.
Changes:
- Add src/prompts/ directory with 7 prompt files (wizard, AutoRun, etc.)
- Add index.ts for central exports using Vite ?raw imports
- Add esbuild plugin in build-cli.mjs to support ?raw imports for CLI
- Update wizardPrompts.ts and phaseGenerator.ts to use central substitution
- Update CLAUDE.md documentation with new prompt location references
- Add TypeScript declaration for *.md?raw imports in global.d.ts
Claude ID: 38553613-f82f-4ce1-973e-fa80d42af3da
Maestro ID: b9bc0d08-5be2-4fdf-93cd-5618a8d53b35
