You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Generate vocabulary hints for voice dictation in this project.
Scan the codebase and extract terms that a speech-to-text model would likely misrecognize — class names, function names, variable names, technical jargon, library names, acronyms, and project-specific vocabulary.
Guidelines:
- One term per line
- Lines starting with # are comments
- Include: class/struct/type names, important function names, library/tool names, domain-specific terms, proper nouns (e.g. "Claude", "Kubernetes", "PostgreSQL")
- Skip: common English words, single-letter variables, auto-generated names
- Keep it under 100 terms — focus on words the model is most likely to get wrong
Create the directory `.dictate-hints.d/` if it doesn't exist, then write to `.dictate-hints.d/project.hints`. Do not touch other files in that directory — they may contain manually curated hints.