Skip to content

About

Extensible plugin ecosystem for Caster voice programming featuring reactive PyQt HUD overlays, taskbar telemetry, and ADCE context engines

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Caster Plugins Repository

Dedicated distribution repository for optional Caster plugins, visual overlays, and external bridges.

Taskbar HUD & ADCE Live Telemetry

Live Telemetry & Context In Action: Combining taskbar_hud with the adce (Active Desktop Context Engine) plugin projects real-time speech hypotheses, executed commands, microphone states, and sub-window semantic focus zones directly into the Windows 11 Taskbar system tray.


Overview

This repository hosts standalone plugins for the Caster voice programming framework. By separating advanced integrations from Caster Core, the core repository remains minimal and focused strictly on grammar execution.

Available Plugins

Plugin Version Description Target Environment
taskbar_hud 1.1.0 Named Pipe bridge projecting real-time speech telemetry directly into the Windows 11 Taskbar via Caster Taskbar HUD with event-driven Win32 hooks and two-phase context resolution. Windows 11 Taskbar
adce 1.0.0 Active Desktop Context Engine SSE client providing sub-millisecond focus predicates for integrated terminals. Local HTTP/SSE Port 8424
plugin_manager 1.0.0 Extensible plugin manager GUI and headless registry engine with schema-driven configuration and companion voice rules. Desktop Pop-up Dialog
dynamic_cursor 1.0.0 Event-driven bridge projecting Caster mic states and ADCE semantic focus badges onto the Windows hardware cursor via dynamic-cursor. Windows Hardware Cursor

Shared Architecture (plugins/common)

All visual HUD plugins share a common foundation in plugins/common:

  • Unified Context Resolver (context_resolver.py): Prioritizes active in-memory Caster runtime data (nexus._grammar_manager._managed_rules) with AST fallbacks for offline testing. Implements two-phase evaluation (Phase 1: executable/title filtering; Phase 2: safe FuncContext execution).
  • Shell Overlay Exclusions: Automatically suppresses Windows Shell infrastructure windows (Alt+Tab overlays, taskbars, desktop backdrops) to prevent false-positive rule matches.
  • Plugin Lifecycle Fallbacks (plugin_base.py): Provides standard base classes ensuring seamless execution across varied environments.

Installation Workflow

Users install plugins into their personal user directory via the Caster plugin CLI:

py -3.10 -m castervoice.bin.plugin_cli list
py -3.10 -m castervoice.bin.plugin_cli install taskbar_hud

Plugins install into %LOCALAPPDATA%\caster\caster_user_content\plugins\<name>\ and activate automatically in settings.toml.

Contributing a Plugin

Each plugin directory must contain:

  1. metadata.toml: Declarative manifest defining name, version, description, and dependencies.
  2. plugin.py: A subclass of PluginBase implementing initialize(), start(), and stop().
  3. __init__.py: Package export file.

About

Extensible plugin ecosystem for Caster voice programming featuring reactive PyQt HUD overlays, taskbar telemetry, and ADCE context engines

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages