Des transcriptions de podcasts aux publications

Les consignes originales en anglais ci-dessous sont aussi celles du skill copié ou téléchargé.

# Podcast transcripts → LinkedIn posts

You have the "lead-scorer" MCP server connected (Lead Scorer CRM — endpoint https://mcp.lead-scorer.com/mcp, authenticated with Lead Scorer OAuth). Use its tools for every read and write. Discover resource IDs with the available list/search tools; never guess or probe sequential IDs, and ask me when no discovery tool exists. Never invent data: if a tool result is empty, say so. An API key is only a manual fallback for clients without OAuth support.

## Goal
Turn newly stored podcast transcripts into LinkedIn drafts built on what was ACTUALLY said.

## Steps
1. **Discover.** `list_content_items` with item_type "podcast_episode", added today. Pick the 1-3 most relevant for my audience.
2. **READ THE FULL TRANSCRIPT** via `get_content_item` (include_transcript). NEVER write from the episode title: titles are marketing. (Real case: an episode titled "Founder-led sales" was actually a podcast-marketing retrospective — a title-based post would have been fiction.)
3. **Extract the operator insight** — the lesson a practitioner takes away, not an episode summary. If a transcript has no exploitable angle, skip it and say why.
4. **Write 1-2 posts per kept episode.** Name the guest and their role; translate their best formulas as near-verbatim quotes. Same voice rules as my other content skills: fact-first hook, one line of my lived experience, "so what?" test, no emoji, no em-dash, sincere closing question.
5. **Store.** `create_content_post`, status "draft", external_id "post:linkedin/<slug>".

## Hard rules
- A quote must exist in the transcript. Paraphrase honestly or quote exactly; never improve someone's words.
- Draft only; I review everything.