Add Decktation to Plugin Store - #1126
Open
silverfoxy wants to merge 1 commit into
Open
silverfoxy wants to merge 1 commit into
silverfoxy wants to merge 1 commit into
Conversation
Issues Found
Next Steps
Thank you for your contribution! If you need any help, please reach out on our Discord server. ❤️ |
silverfoxy
requested a deployment
to
testing_env
September 12, 2026 23:38 — with
GitHub Actions
Waiting
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add Decktation to Plugin Store
Decktation is a push to talk voice to text plugin for Steam Deck. It records speech using a configurable button combo, transcribes it locally via faster-whisper model and types the result into the active game or application.
The plugin includes presets for World of Warcraft, Guild Wars 2, and a generic text input. The game presets can open chat, recognize channel names (party -> /p) and send the message to the appropriate channels. Users can customize the button combo, select different model sizes (larger models for better accuracy but slower transcription), different languages (as supported by the Whisper model), and a few miscellaneous options such as typing the text but waiting for the user to press enter. Or showing the preview text as a toast message allowing the user to cancel the message before sending it.
The plugin bundles its Python dependencies and ydotool helper (to type) so users do not need sudo or to manually install any system packages (the original version required manual installation in desktop mode).
Link to repo: https://github.com/silverfoxy/decktation
Reddit post: https://www.reddit.com/r/SteamDeck/comments/1rcmjxb/decktation_deckyloader_voice_to_text_plugin_for/
Video: https://www.youtube.com/watch?v=qGRD6njtbNI
Blog post (Installation has been streamlined since writing this blog post): https://imsilverfoxy.com/posts/decktation-voice-to-text-for-wow-on-steamdeck/
Task Checklist
Developer
I have used Claude code and Codex to research existing approaches for ASR (Automatic Speech Recognition), benchmark CPU-based models that can run on Steam deck, PoC and test my plugin, and fix bugs. I have actively maintained and supported Decktation in the past 6 months through feature additions based on requests from Reddit users as well as bug fixes as evident by the Reddit threads activity. I fully designed the architecture and inner workings of my plugin and I am capable and willing to support the plugin for the foreseeable future.
Plugin
Currently decky store lists decky-dictation, a voice to text plugin that uses the Vosk transcription model and has not been updated for over 2 years. Although Decky Dictation (https://github.com/antiquitte/decky-dictation) demonstrated the basic idea of speech to text on Steam Deck, Decktation is an independent implementation rather than a fork or direct continuation.
Decky Dictation and its predecessor were the main motivation for me to build Decktation. But I changed the underlying speech recognition model to a modern ASR model (Whisper), and added numerous features and configurations. Decktation owns the complete speech pipeline instead of acting as a launcher for an external dictation program. Decktation records and normalizes audio itself, runs faster-whisper through the optimized CTranslate2 engine using CPU int8 inference, and manages transcription results before deciding where and how to type them. As a result I would not consider them the same plugin.
Backend
Community
I will update this section once I test existing plugins. As a side note, the current state of documentation around testing is confusing. The repo's readme links to the SDH tracker dashboard that shows no plugins available for test. The test channel inside decky loader lists plugins that are not among the open PRs on Github. My understanding is that Discord is the source of truth.
Testing