Use transcripts published in the RSS feed (podcast:transcript) before paid transcription #348

Problem. Users want transcripts like Readwise Reader or Snipd ("import a whole transcript (synced or time coded to the audio would be a bonus)", https://forum.obsidian.md/t/108426/12). PodNotes only transcribes through OpenAI or Deepgram with the user's key, and six 2024-2025 issues show that path failing on credit, rate limits, m4a chunking and local files (#102, #111, #116, #117, #118). Many feeds publish <podcast:transcript> (VTT, SRT, JSON) for free, and the parser already reads podcast:chapters (src/parser/feedDocumentParser.ts) but not transcripts. #91 asked for this from Apple and was closed after API transcription shipped.

Proposal. Parse podcast:transcript. When present, the Transcribe command writes it with timestamps (as links when the template asks), with no API key.

Acceptance. For a feed with a VTT transcript, Transcribe produces a timestamped transcript note offline from the API. A parser test uses a real feed fixture.

Found by the PodNotes user-feedback research (Reddit, Obsidian forum, YouTube, GitHub history).