osagent 0.2.78 → 0.2.80

package/README.md

@@ -1,4 +1,4 @@
-# OS Agent
+# osagent
 
 An AI-powered autonomous coding assistant CLI that runs locally with Ollama or connects to cloud AI providers. Built for developers who want full control over their AI coding assistant.
 
@@ -93,7 +93,7 @@ npm install -g osagent
    ```bash
    ollama pull qwen2.5-coder:32b
    ```
-3. Run OS Agent:
+3. Run osagent:
    ```bash
    osagent
    ```
@@ -105,7 +105,7 @@ npm install -g osagent
    ```bash
    export OLLAMA_API_KEY=your-api-key
    ```
-3. Run OS Agent:
+3. Run osagent:
    ```bash
    osagent
    ```
@@ -117,7 +117,7 @@ npm install -g osagent
    ```bash
    export ANTHROPIC_API_KEY=your-api-key
    ```
-3. Run OS Agent:
+3. Run osagent:
    ```bash
    osagent --auth anthropic
    ```
@@ -129,7 +129,7 @@ npm install -g osagent
    ```bash
    export GROQ_API_KEY=your-api-key
    ```
-3. Run OS Agent:
+3. Run osagent:
    ```bash
    osagent --auth groq
    ```
@@ -175,7 +175,7 @@ Once in the interactive mode, you can use these commands:
 - `/help` - Show available commands
 - `/about` - Show version and system info
 - `/settings` - View and edit settings
-- `/quit` or `/q` - Exit OS Agent
+- `/quit` or `/q` - Exit osagent
 
 #### Model & AI
 - `/model` - Switch AI model (opens model selector)
@@ -184,7 +184,7 @@ Once in the interactive mode, you can use these commands:
 - `/model <id>` - Switch to a specific model
 - `/provider` - Switch AI provider (Ollama, Anthropic, GROQ, OpenAI)
 - `/agents` - Manage custom agents
-- `/consult` - Toggle consultation mode (AI asks clarifying questions)
+- `/consult` - Manage consultation mode (AI asks clarifying questions)
 
 #### Conversation
 - `/continue` - Resume the most recent conversation
@@ -216,7 +216,7 @@ Once in the interactive mode, you can use these commands:
 
 ### Status Bar
 
-While OS Agent works, you'll see a real-time status bar showing:
+While osagent works, you'll see a real-time status bar showing:
 - Current activity with spinner
 - Token usage (input ↑ / output ↓)
 - Elapsed time
@@ -233,16 +233,18 @@ While OS Agent works, you'll see a real-time status bar showing:
 
 ### Consultation Mode
 
-OS Agent can ask you clarifying questions as it works to better understand your requirements:
+osagent can ask you clarifying questions as it works to better understand your requirements:
 
 ```bash
-/consult on        # Active mode - questions shown while agent works
+/consult active    # Active mode - questions shown while agent works
 /consult blocking  # Blocking mode - high-priority questions pause the agent
-/consult off       # Disable consultation
+/consult status    # Show current consultation mode and stats
 /consult context   # View collected context
+/consult clear     # Clear collected consultation context
+/consult demo      # Run a demonstration of consultation features
 ```
 
-Questions appear in a dedicated panel below the status bar during agent operations.
+Consultation is always enabled. Use `/consult active` or `/consult blocking` to switch between modes. Questions appear in a dedicated panel below the status bar during agent operations.
 
 ### Vector Memory (Optional)
 
@@ -258,18 +260,18 @@ ollama pull nomic-embed-text
 
 ### Multi-Agent Orchestration
 
-OS Agent features an intelligent orchestration system that automatically:
+osagent features an intelligent orchestration system that automatically:
 
 - **Detects skills** in your prompts (debugging, testing, refactoring, etc.)
 - **Selects the best agent** for each task based on capabilities
 - **Dispatches execution** to specialized subagents
 - **Tracks progress** with real-time status updates
 
-The orchestrator runs transparently - just describe what you want and OS Agent handles the rest.
+The orchestrator runs transparently - just describe what you want and osagent handles the rest.
 
 ### Episodic Memory
 
-OS Agent learns from your sessions:
+osagent learns from your sessions:
 
 - **Records episodes** of your interactions automatically
 - **Tracks actions** including tools used and files modified
@@ -280,7 +282,7 @@ Episodes are stored locally in `~/.osagent/episodes/`.
 
 ### Configuration
 
-OS Agent stores its configuration in `~/.osagent/` (created automatically on first run):
+osagent stores its configuration in `~/.osagent/` (created automatically on first run):
 
 ```
 ~/.osagent/
@@ -306,7 +308,7 @@ model: qwen2.5-coder:32b
 
 ### Project Memory
 
-Create an `OSAGENT.md` file in your project root to give OS Agent context:
+Create an `OSAGENT.md` file in your project root to give osagent context:
 
 ```bash
 osagent /init
@@ -355,7 +357,7 @@ npm run lint
 
 ## Architecture
 
-OS Agent is a monorepo with the following packages:
+osagent is a monorepo with the following packages:
 
 ```
 packages/
@@ -378,8 +380,6 @@ packages/
 ### During Operations
 - `Ctrl+O` - Expand/collapse output
 - `Ctrl+Y` - Toggle todo list visibility
-- `Ctrl+E` - Toggle agent detail level
-- `Ctrl+F` - Focus embedded shell
 
 ### Text Editing
 - `Ctrl+L` - Clear screen
@@ -391,135 +391,10 @@ packages/
 
 ## Documentation
 
-For detailed architecture documentation, see [SYSTEM_ARCHITECTURE.md](SYSTEM_ARCHITECTURE.md). This includes:
-- Complete request flow diagrams
-- Agent system and tier structure
-- Skill detection and auto-dispatch
-- Tool system and execution flow
-- Memory systems (episodic + vector)
-- Context provider hierarchy
-- Integration points and key files
-
-## Recent Updates
-
-### v0.2.6 - Smart Agent Selection
-- **Intelligent Orchestration**: Generic "build X" requests now use Master Orchestrator
-- **Confidence Thresholds**: Specialists only selected when confidence > 0.6
-- **New Project Detection**: Automatically routes new project requests to Orchestrator
-- **Tech Keyword Detection**: Only uses specialists when specific tech is mentioned
-
-### v0.2.5 - Comprehensive Orchestrator
-- **Expert-Level Orchestrator**: Full tool knowledge with decision framework
-- **Problem-Solving Framework**: Identify, analyze, plan, implement, verify phases
-- **Feature Building Framework**: Understand, plan, build, verify, complete phases
-- **Action-Oriented Prompts**: Models now explain while acting, not just talking
-
-### v0.2.4 - Agentic Enforcement
-- **Critical Operating Principle**: Models must ACT, not just describe actions
-- **Tool Call Requirements**: Every action statement must include tool calls
-- **Improved System Prompts**: Clearer instructions for tool usage
-
-### v0.2.3 - API Key Fixes
-- Fixed paste handling in API key input (bracketed paste mode)
-- Added `/provider reset` command to clear stored API keys
-- Removed sudo from doctor command recommendations
-
-### v0.2.2 - Ollama Cloud Fix
-- Fixed Ollama Cloud API key not being stored in environment
-- API key now properly persists for session after entry
-
-### v0.2.1 - Bundle Fix
-- Fixed version number not showing correctly in bundled CLI
-- Use `npm run bundle` instead of `npm run build` for releases
-
-### v0.2.0 - Architecture & Orchestration Improvements
-- **Roadmap System**: New persistent project tracking system (`roadmap` tool for agents)
-  - Track work items with priorities, dependencies, and subtasks
-  - Stored in `.osagent/roadmap.json` - persists across sessions
-  - Actions: view, add, update, next, summary, note
-- **Consultation Integration**: Enhanced orchestrator-consultation bridge
-  - Generates clarifying questions when prompts are ambiguous
-  - Context-aware question generation based on skill detection
-  - Collects enhanced context for better agent selection
-- **Codebase Cleanup**: Removed orphaned code and redundant hooks
-  - Deleted unused memory command hooks
-  - Consolidated RoadmapContext into core service
-  - Updated architecture documentation
-- **Documentation**: Added comprehensive SYSTEM_ARCHITECTURE.md
-  - Complete system flow documentation
-  - Agent tiers and selection rules
-  - Skill detection process
-  - Tool registration patterns
-
-### v0.1.87 - Test Fixes & Documentation
-- Fixed all test failures (3170/3170 tests now passing)
-- Updated naming consistency across codebase
-- Fixed vision model detection in DashScope provider
-- Added gopd dependency troubleshooting documentation
-- Improved MCP client transport handling
-
-### v0.1.86 - Auto Project Context
-- Automatic project context injection on first prompt
-- When no OSAGENT.md exists, AI auto-explores project structure before responding
-- No manual `/init` required for casual usage
-
-### v0.1.85 - Security & Stability Fixes
-- Timer now only counts active work time (stops when idle)
-- Fixed command injection vulnerability in git co-author feature
-- Fixed symlink bypass in path validation
-- Fixed VimModeContext re-render loop
-
-### v0.1.83 - Enhanced Doctor Command
-- Added enhanced installation diagnostics to `/doctor`
-- Shows installation method (npm-global, homebrew, native, source)
-- Displays binary path and invoked path
-- Shows auto-updates status and search/ripgrep status
-- Detects and warns about multiple installations
-- New "All systems go" summary style
-
-### v0.1.81 - Stability & Auto-Initialization
-- Fixed "Maximum update depth exceeded" infinite loop (3800+ errors)
-- Added automatic `~/.osagent/` directory initialization on startup
-- Creates required directories: agents, commands, prompts, skills, episodes
-- Auth credentials now properly saved when switching providers
-
-### v0.1.80 - Auth Provider Fix
-- Fixed auth configuration mismatch when switching providers
-- Ensures baseUrl and model are saved correctly to settings
-
-### v0.1.79 - IDE Terminal Compatibility
-- Migrated all dialogs from custom `useKeypress` to Ink's `useInput` hook
-- Fixes arrow key issues in VS Code, Cursor, and other IDE terminals
-- Arrow keys no longer print escape sequences (`^[[A^[[B`) in dialogs
-- Improved compatibility with IDE terminals that don't support raw mode
-
-### v0.1.66 - Focus Management
-- New `useFocusedKeypress` hook for enforced focus management
-- Prevents keypress conflicts between dialogs and main UI
-- Required `isFocused` parameter catches bugs at compile time
-
-### v0.1.65 - Input Routing Fix
-- Fixed keyboard input routing issues with selectors
-- Arrow keys and Enter now properly isolated to focused component
-- No more keypress "leakage" between dialogs
-
-### v0.1.64 - UI Improvements
-- Tree-view style tool display with `├─` `└─` connectors
-- Expandable results with `Ctrl+O`
-- Token counts and duration stats
-- Sleeker ASCII art header
-- Minimal init banner
-
-### v0.1.63 - Stability Fix
-- Fixed infinite loop on startup (Maximum update depth exceeded)
-- Improved React effect dependency management
-
-### v0.1.59 - Terminal Stability
-- Fixed EIO error crash on stdin
-- Graceful handling of TTY disconnects
-- Stable arrow key navigation in selectors
-
-See [KNOWN_ISSUES.md](docs/KNOWN_ISSUES.md) for detailed issue tracking.
+- [ARCHITECTURE.md](ARCHITECTURE.md) - System architecture, message flow, content generators, and tool system
+- [CHANGELOG.md](CHANGELOG.md) - Full version history and release notes
+- [docs/](docs/) - Detailed documentation on CLI, tools, core APIs, and features
+- [docs/reference/](docs/reference/) - System architecture deep-dive, UI/UX flow, tech stack
 
 ## License