Skip to content

hardhackerlabs/podwise-cli / skills

podwise

Flagged

Podcast knowledge workflows powered by Podwise CLI: search podcasts and episodes by keyword, monitor followed shows for new releases, find popular episodes, ask questions and extract insights from transcript content, process Podwise episode URLs, YouTube videos, Xiaoyuzhou links, and local audio or video files to retrieve transcripts, summaries, chapters, Q&A, mind maps, highlights, and keywords — plus catch up on your backlog, refine your listening taste, generate weekly recaps, export episode notes to PKM tools, research topics across podcasts, debate episode ideas, and generate language learning cards. Use when the user wants to find, summarize, transcribe, or extract insights from any podcast or audio content, or manage their listening library.

workflows/episode-notes.md

View on GitHub
# Podwise Episode Notes

Use this skill to turn a single processed episode into a well-structured, portable note. It pulls every available AI artifact from Podwise and assembles them into one document shaped for long-term reference — not just a quick read.

## Goals

1. Verify that `podwise` is installed and configured.
2. Identify the target episode from the user's input.
3. Ensure the episode is processed; if not, ask for confirmation before processing.
4. Fetch all available AI artifacts for the episode.
5. Assemble a structured markdown note.
6. Write the note to a file and guide the user on how to import it into their PKM tool.

## Step 1: Check the Environment

```bash
podwise --help
podwise config show
```

If `podwise` is not installed or the API key is missing, stop and follow [references/installation.md](references/installation.md) before continuing.

## Step 2: Load the Listener Taste

Look for `taste.md` in the current working directory.

- If found, read the **PKM tool** and **Output Preferences** fields silently. Use these to shape the note format and to give accurate export instructions in Step 7.
- If not found, proceed with default formatting and generic export instructions.

## Step 3: Identify the Target Episode

The user may provide the episode in several ways:

- A Podwise episode URL: `https://podwise.ai/dashboard/episodes/{id}`
- A YouTube URL: `https://www.youtube.com/watch?v=...`
- A Xiaoyuzhou URL: `https://www.xiaoyuzhoufm.com/episode/...`
- A local audio or video file path: `./recording.mp3`
- An episode title or keyword — in this case, search first:

```bash
podwise search episode "{title or keyword}" --limit 5 --json
```

Present the results and ask the user to confirm which episode they want before continuing.

If the user provides a YouTube, Xiaoyuzhou, or local file input, that content must be processed before artifacts can be fetched. Move to Step 4 immediately.

## Step 4: Ensure the Episode Is Processed

For a Podwise episode URL, attempt to fetch the summary to check if processing is complete:

```bash
podwise get summary {episode-url}
```

- If this succeeds, the episode is already processed. Skip to Step 5.
- If this returns an error indicating the episode is not yet processed, tell the user and ask for confirmation before running `process`:

> "This episode hasn't been processed yet. Processing will use one credit from your Podwise quota. Proceed?"

Only run `process` after explicit confirmation:

```bash
podwise process {episode-url}
```

For YouTube, Xiaoyuzhou URLs, or local files, always ask for confirmation before processing — these always consume quota:

```bash
podwise process {youtube-or-xiaoyuzhou-url}
podwise process {local-file-path} --title "{title}"
```

`podwise process` waits for completion before exiting. Once it returns successfully, the Podwise episode URL will be available in its output — use that URL for all subsequent `get` commands.

## Step 5: Fetch All AI Artifacts

Run all `get` commands for the episode:

```bash
podwise get summary {episode-url}
podwise get chapters {episode-url}
podwise get highlights {episode-url}
podwise get qa {episode-url}
podwise get keywords {episode-url}
podwise get mindmap {episode-url}
```

Run these in sequence. If any individual command fails, mark that artifact as unavailable in the note rather than stopping the whole skill.

Optionally fetch the transcript if the user specifically requested it or if their taste profile indicates they prefer full transcripts:

```bash
podwise get transcript {episode-url}
```

Do not include the full transcript in the note by default — it is too long for most PKM use cases. Offer it as a separate file if the user wants it.

## Step 6: Assemble the Note

Combine all fetched artifacts into a single markdown document using this structure:

---

```markdown
# {Episode Title}

**Podcast**: {Podcast Name}
**Published**: {publication date}
**Source**: {episode-url}
**Processed**: {date note was generated}

---

## Summary

{Content from `get summary`}

---

## Chapters

{Content from `get chapters`, formatted as a numbered list with timestamps if available}

---

## Highlights

{Content from `get highlights`, formatted as a bulleted list}

---

## Q&A

{Content from `get qa`, formatted as question/answer pairs}

---

## Keywords

{Content from `get keywords`, formatted as a comma-separated inline list or tag-style list depending on user's PKM tool}

---

## Mind Map

{Content from `get mindmap`, rendered as a nested bullet list if the raw output is a tree structure}

---

_Note generated by podwise-episode-notes · {date}_
_Source: {episode-url}_
```

---

Rules for assembling the note:

- Omit any section where the artifact was unavailable — do not show empty section headers.
- Keep the Summary section as returned by `get summary` without paraphrasing or shortening it.
- Highlights should remain verbatim from `get highlights` — do not rewrite them.
- If `get mindmap` returns structured data rather than plain text, convert it to a nested markdown bullet list before inserting it.
- If the user's taste profile specifies a preferred output format (bullet points vs prose), apply it only to synthesized sections — do not alter verbatim artifacts.

## Step 7: Write the File and Guide Export

Name the file using the pattern: `{podcast-name}-{episode-slug}-notes.md`

Slugify by lowercasing, replacing spaces with hyphens, and stripping special characters. Example:
`lex-fridman-podcast-elon-musk-notes.md`

Write the file to the current working directory unless the user specifies another path.

Then tell the user where the file was saved and provide import instructions based on their PKM tool from the taste profile:

**Notion**
> Drag the `.md` file into any Notion page, or use Notion's "Import" option (File → Import → Markdown & CSV). The headings will map to Notion's heading blocks automatically.

**Obsidian**
> Move the file into your vault folder. It will appear immediately in Obsidian's file explorer. All internal `[[links]]` you add later will connect to it like any other note.

**Logseq**
> Place the file in your Logseq `pages/` directory. Logseq will index it on next open. The `## Highlights` section works well as block references.

**Readwise**
> Use Readwise's manual highlight import or the Readwise Reader upload feature. Note that Readwise is optimised for highlights rather than full notes — consider importing just the Highlights section.

**No PKM tool in taste profile / unknown**
> The note is saved as a plain markdown file at `{path}`. You can open it in any text editor, import it into most note-taking apps, or keep it as a standalone reference.

## Common Failure Cases

- If the episode search returns no results, ask the user to try a different title, keyword, or to paste the episode URL directly.
- If `process` fails due to an unsupported file format, stop and list the supported extensions: `.mp3 .wav .m4a .mp4 .m4v .mov .webm`.
- If the user provides a URL that is neither a valid Podwise, YouTube, Xiaoyuzhou, nor local path, tell them the URL format is not recognised and ask for a supported input.
- If all `get` commands fail after a successful `process`, tell the user that processing may still be completing and to try again in a few minutes.
- If a note file with the same name already exists, ask the user whether to overwrite it or save with a timestamp suffix.

## Output Contract

Produce exactly one markdown note file per episode.

The note must include at minimum: Summary, Highlights, and the source URL. All other sections are included when available.

The transcript is never included in the note by default — offer it separately only if requested.

Always confirm the file path after writing and always provide PKM import instructions.
add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope.Stars: 114kUpdated Claude CodeCodexVerifiedInstall
fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks.Stars: 114kUpdated Claude CodeCodexVerifiedInstall