do-voice-recording

# /do-voice-recording — Text → Spoken Audio The single, simple text-to-speech surface for the whole system. Hand it text, get back a spoken-audio file (OGG/Opus). Every other skill that needs synthesis (`/do-presentation` voiceovers, `/do-debrief` voice notes) defers to this rather than reimplementing it. ## Repo Context Probe If `.claude/skill-context/do-voice-recording.md` exists, read it and honor its declarations; otherwise use the generic defaults described below. The context file is where a repo declares the actual TTS CLI this skill drives: how to resolve the binary portably regardless of cwd, the synthesize command and its flags, the voice catalog, and how to deliver the result (e.g. as a chat voice note). When the file is absent (the common case in a foreign repo), follow the generic baseline below. ## Generic baseline — TTS requires a repo-provided CLI Text-to-speech is not a capability the bare environment provides — it needs a synthesis engine. This skill does not bundle one; it drives whatever TTS CLI the repo supplies and documents in its context file. - **Context file present** → resolve and invoke the declared TTS command exactly as it specifies, passing the user's text and any `--voice` / output options. It returns (or names) the produced audio file path. - **Context file absent** → the synthesis dependency is unavailable in this repo. Tell the user that text-to-speech requires a repo-provided CLI which this repo does not declare, and stop gracefully. ...