Voice Report Generator

Problem Addressed

In office workflows, many topic briefs require manual steps: gathering sources, structuring a written report, and converting it into playable audio. This skill turns that sequence into an executable pipeline for topic input, report generation, speech synthesis, and link delivery, reducing manual handoffs.

How It Works

• Report generation: It searches for factual information and produces a plain-text report of at least 500 words, structured as objective facts followed by subjective evaluation; the evaluation should cover current state, causes, trends, and personal opinion.
• Speech synthesis: It splits the text into natural 100–120 word segments, preferring the Edge TTS hosted API; on timeout or failure, it falls back to node-edge-tts and generates MP3 files with Xiaoxiao Neural.
• Delivery: It merges audio files with ffmpeg, uploads them using upload.py, and returns an HTTP audio link for direct playback.

Boundaries

The skill is best suited for quick sharing, spoken summaries, and audio briefs. It outputs a fixed single-paragraph report and an HTTP MP3 link. If HTTPS links, private intranet storage, editable document formats, or multi-round fact checking are required, add human review before delivery.