ttslisted

# tts — document to speech Turns a document into an MP3: (1) rewrite it into speech-friendly prose, (2) render with a neural TTS engine (`edge-tts` by default). TTS engines mangle code, symbols, lists, and markdown, so the rewrite is where the value is. ## Requirements - Python 3.9+ to run this skill's own scripts. - A neural TTS engine, verified once via the bootstrap step below. `edge-tts` is the default this skill is built around; see `resources/bootstrap.py` for other free and paid options. What this skill executes, sends over the network, and writes to disk is disclosed in [README.md, "Permissions and network egress"](README.md#permissions-and-network-egress). ## Bootstrap Check whether `<SKILL_DIR>/.bootstrap-verified` exists. - **If it exists**, an engine is already verified — skip straight to Inputs. (`python3 <SKILL_DIR>/resources/bootstrap.py --verified` prints which one, if you need to mention it.) - **If it doesn't**, read and follow `<SKILL_DIR>/resources/bootstrap-instructions.md` in full. ## Inputs 1. Confirm the **Source file** — a path passed with the request. If none was given, ask for one. Accept `.md`, `.txt`, or any plain-text file. (For PDFs, extract text first, e.g. with `pdftotext`.) 2. Always ask for **Output folder** with AskUserQuestion (single-select). Options: - Current directory - Source document's folder — include this option ONLY if it differs from the current directory - Other (let the user type a path) The output filename i