Skip to content
Agentmart
Submit a tool

podcast

Produce an audio podcast from source material (doc, text, link, topic) via Google NotebookLM's MCP server.

Skill by smlfg

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/smlfg-podcast

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Produce an audio podcast from source material (doc, text, link, topic) via Google NotebookLM's MCP server. Use when the user asks for a podcast, an audio overview, or "turn this into audio".

Use cases: documents, backend, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/podcast

What's inside

Skill name
podcast
Description
Produce an audio podcast from source material (doc, text, link, topic) via Google NotebookLM's MCP server. Use when the user asks for a podcast, an audio overview, or "turn this into audio".
License
MIT
Instructions
43 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Podcast via NotebookLM

When the user wants a podcast, **always use NotebookLM** (via a NotebookLM MCP server, tools named like `notebook_create`, `source_add`, `studio_create`, `studio_status`, `download_artifact`). Never hand-roll audio.

## Before the run

**Check auth** (e.g. `server_info`). If not configured, ask the user to run the CLI login (e.g. `nlm login`) in a terminal, then continue.

## Steps

1. **Clarify the source.** Three cases:
   - Existing doc/file → use the path. **If it lives in a project folder, add the whole relevant folder** (all .md/.txt/.pdf) as sources, not just one file; more context makes NotebookLM better.
   - Topic without text → first write a short, speakable source text (~700–1000 words, `##` sections, honest, no marketing). For larger content, delegate the writing to a cheaper model.
   - Link / YouTube → URL source.
2. **Create a notebook**: `notebook_create(title=...)`, keep the `notebook_id`.
3. **Add sources**: `source_add(notebook_id, source_type=file|text|url, ..., wait=true)`.
4. **Generate audio**: `studio_create(notebook_id, artifact_type="audio", language="<lang>", custom_prompt="<goal, tone, honesty notes>", confirm=true)`. Default format is the two-voice deep dive; shorten via the length option.
5. **Poll** `studio_status(notebook_id)` until ready. Generation takes minutes; don't block the main session, check at intervals.
6. **Download**: `download_artifact(notebook_id, artifact_type="audio", output_path="<project>/<name>.m4a", wait=true)`.
7. **Report** the audio path and notebook URL.

## Principles

- **Delta, not recap.** The listener usually knows their own plans and documents. Always instruct: "The listener already knows the context; no retelling of the starting point; focus on what's new, decisions, conflicts, reasons, next steps." For recap podcasts choose a shorter length; 30 minutes is usually too long.
- **Honesty in content**: don't claim features or capabilities that don't exist (especially for onboarding/product podcasts).
- **Language**: use the user's preferred language by default.
- Clarify tone, length and focus briefly if unclear; don't guess.

## Example

```text
Source: docs/architecture/*.md (6 files)
custom_prompt: "Listener built this system. Skip the overview. Focus on the two
decisions made this week (queue → event log; drop the cache) and their trade-offs.
Max ~12 minutes. Don't claim the migration is finished — it isn't."
→ ./podcasts/arch-decisions-week-41.m4a + notebook URL
```

## Use case

You wrote a long design doc and want to absorb the decisions on a walk. Ask for a podcast: the agent loads the whole docs folder into NotebookLM, steers it toward "what changed and why", and drops an m4a next to the doc.

From smlfg/HAI-skills-registry, MIT.

More from smlfg/HAI-skills-registry

Compare and alternatives

    Featured in

    Signals

    Stars
    0
    Forks
    0
    Open issues
    0
    Last commit (this folder)
    Oct 15, 2026 (13 days ago)

    From GitHub (fetched Oct 24, 2026).

    Details

    License
    MIT
    First seen
    Oct 24, 2026
    Last checked
    Oct 24, 2026

    More from smlfg

    Cite this page

    podcast skill. Agent Mart. https://agentmart.dev/skills/smlfg-podcast (last updated 2026-10-24)