Generating Podcast Script

Podcast Script Generation

Produces a two-host, NotebookLM-style conversational episode: a readable script plus a multi-voice SSML document ready for Azure Text-to-Speech.

Step 1 — Gather inputs

Ask only for what is missing. Do not ask for optional fields that have safe defaults.

All word targets assume ~150 spoken words per minute. Stay within 10 percent of target.

Derive a lowercase-hyphenated <slug> from the source or topic (max 6 words, ASCII only).

Step 2 — Parse the source material

Skip this step when the user gave a bare topic with no source material.

Extract every distinct article or item. For each, capture:

• headline
• publication / source
• date, if present
• the core factual claim
• any figures or quotes
• the "so what"

Merge duplicates covering the same event. Discard boilerplate, footers, disclaimers, legal notices, unsubscribe text, and image captions.

Step 3 — Editorial selection

Rank items by newsworthiness and reader impact. Keep the top 4 to 6 for full treatment (fewer for short, more for long). Group the remainder into one fast rapid fire segment. If the material has a dominant theme, lead with it and thread it through the episode.

Step 4 — Cast

Two hosts, always the same personalities, always distinct voices:

• NOVA — voice en-US-AvaMultilingualNeural. Lead host. Warm, curious, quick. Drives the agenda, asks the question the listener is thinking, reacts out loud, reframes jargon into plain language. Slightly faster cadence.
• MILES — voice en-US-AndrewMultilingualNeural. Analyst. Calm, dry, precise. Supplies context, numbers, caveats, second-order implications. Slightly slower, lower pitch. Occasionally pushes back on Nova.

Neither host is a narrator. They talk to each other, not to the microphone.

Step 5 — Episode structure

1. Cold open (15–20s) — Nova opens on the single most striking fact or tension. No "welcome to the podcast", no channel branding, no music cues.
2. Agenda tease (10s) — Miles lays out what they'll cover, casually.
3. Story segments (60–90s each) — one item per segment. Pattern: hook → the facts → why it matters → a short exchange of interpretation → handoff line into the next story. Vary who leads each segment.
4. Rapid fire (45s) — alternating one-liners on the leftover items, quick tempo.
5. Close (20–30s) — three concrete takeaways split between the two hosts, then a short human sign-off. Alternate whether Nova or Miles ends.

Step 6 — Dialogue style

This is what makes it sound conversational rather than read-aloud.

• Write spoken English, not written English. Contractions everywhere.
• Keep most lines under 30 words. Break long explanations across two or three turns with the other host interjecting.
• Use real conversational connective tissue — "okay so", "right", "wait, back up", "here's the part I didn't expect", "yeah, and that's the thing", "hmm". Roughly one marker every 4 to 5 turns. Never let it become a tic.
• One host regularly asks the naive clarifying question so the other can explain.
• Use one concrete analogy or comparison per complex item.
• Genuine reactions are allowed ("that number is wild"). Invented opinions on people, companies, or politics are not.
• Never read a headline verbatim. Paraphrase it into speech.
• Attribute clearly: "according to the Financial Times", "Reuters is reporting".
• If the source is ambiguous or a claim is unconfirmed, say so on air: "the report is careful to call that unconfirmed".
• No stage directions, no "[laughs]", no speaker labels, no markdown, no emojis, no bullet points, and no URLs in the spoken text.

Step 7 — TTS hygiene

Applies to every word that will be spoken.

• Spell out anything a synthesizer would mangle: "twenty twenty-six" not 2026, "three point two billion dollars" not $3.2B, "about fifteen percent" not ~15%.
• First mention of an acronym: expand it, then use the short form.
• Letter-by-letter acronyms: <say-as interpret-as="characters">API</say-as>.
• Odd proper nouns: <sub alias="phonetic spelling">Name</sub>.
• A non-English name or phrase inside an English line: <lang xml:lang="fr-FR">...</lang>.
• Escape XML entities in all spoken text: & → &, < → <, > → >.
• Never emit smart quotes, em dashes, asterisks, or underscores.

Step 8 — Write the readable script

Write the human-readable transcript to:

/app/created/<slug>_Podcast_Script.txt

This file — and only this file — may carry NOVA: / MILES: speaker labels so a person can follow along. It contains no stage directions and no markdown. The spoken text itself must already satisfy Steps 6 and 7 so it can be lifted into SSML unchanged.

Step 9 — SSML output contract

Produce the SSML document and nothing else — no preamble, no explanation, no code fences, no trailing notes inside the artifact. Write it verbatim to:

/app/created/<slug>_Podcast.ssml

Rules:

• Exactly one root <speak> element with version="1.0", xmlns="http://www.w3.org/2001/10/synthesis", xmlns:mstts="http://www.w3.org/2001/mstts", xml:lang="en-US".
• One <voice> element per conversational turn. Alternate speakers. Never put both hosts inside one <voice> element.
• Vary delivery with <prosody> so it never sounds flat. Baselines: Nova rate="+6%" pitch="+2%", Miles rate="-2%" pitch="-4%". Nudge per line to match the emotion of the sentence.
• Use <mstts:express-as style="..."> where the voice supports it. Preferred styles: chat for banter, friendly for explanation, narration-professional for the factual core of a story, excited sparingly for the cold open. An unsupported style is ignored by the service, so keep styles optional — never structural.
• Pauses: <break time="250ms"/> between turns within a segment, <break time="700ms"/> between segments, <break time="400ms"/> before a punchline or a pivot. Never exceed 900ms.
• Every <break> must sit inside a <voice> element. A <break> placed between <voice> elements — as a direct child of <speak> — is invalid in a multi-voice document and will fail synthesis. Two <voice> elements may sit directly next to each other with nothing between them. To pause between turns, put the break at the end of the preceding turn's text, inside that turn's <prosody>.
• <emphasis level="moderate"> on at most one or two key terms per segment.
• Keep the total document under 40,000 characters.

Shape

<speak version="1.0"
       xmlns="http://www.w3.org/2001/10/synthesis"
       xmlns:mstts="http://www.w3.org/2001/mstts"
       xml:lang="en-US">
  <voice name="en-US-AvaMultilingualNeural">
    <mstts:express-as style="excited">
      <prosody rate="+8%" pitch="+3%">Okay, so the number that stopped me cold
      this morning was forty percent. <break time="300ms"/> Forty percent, in one
      quarter. <break time="250ms"/></prosody>
    </mstts:express-as>
  </voice>
  <voice name="en-US-AndrewMultilingualNeural">
    <mstts:express-as style="chat">
      <prosody rate="-2%" pitch="-4%">Right, and the part everyone's skipping is
      that it's off a very small base. <break time="250ms"/> Context matters
      here. <break time="700ms"/></prosody>
    </mstts:express-as>
  </voice>
</speak>

Note the trailing <break> closing each turn: the 250ms is the gap before the next turn, the 700ms is the longer gap before the next segment. Nothing sits between the two <voice> elements.

Step 10 — Review, then generate audio

Show the user a table summarising each segment (title, one-line description, approximate spoken duration), print both file paths, and ask:

Would you like me to convert this to an audio file?

Only if they say yes:

1. Confirm the ConverttexttospeechwithSSML tool is available on the agent. If it is not, tell the user how to add it and stop.
2. Call ConverttexttospeechwithSSML with the SSML document from Step 9 and outputFormat: riff-24khz-16bit-mono-pcm.
3. Decode the base64 response and save to /app/created/<slug>_Podcast.wav.

If the document is too large for a single call, split it at a segment boundary and synthesize each part separately. When stitching, do NOT concatenate decoded RIFF/WAV bytes; instead, append the audio at the PCM-frame level (e.g., via Python’s wave module) and write a single valid /app/created/<slug>_Podcast.wav with one header.

import base64

with open('<tool_output_file>', 'r') as f:
    content = f.read().strip()

audio_bytes = base64.b64decode(content)

with open('/app/created/<slug>_Podcast.wav', 'wb') as f:
    f.write(audio_bytes)

Step 11 — Final report

Always end with:

Run this — do not improvise

This capability's deterministic implementation is a RAPP single-file agent, linked beside this file as generating_podcast_script_agent.py and embedded as the fenced Python below (sha256 efb84e566fdfffb0…; a byte-exact copy is also vaulted in the capsule comment at the end of this file). On a host with sandbox execution, run the linked file directly — if it is missing, write the fence contents verbatim to generating_podcast_script_agent.py first:

python3 generating_podcast_script_agent.py '{"key": "value"}'      # arguments as one JSON object
echo '{"key": "value"}' | python3 generating_podcast_script_agent.py   # or on stdin
python3 generating_podcast_script_agent.py --tool                      # emit the JSON tool contract

Treat stdout as a tool result. If it reports missing or unresolved inputs, stop and collect them. If it returns steps, execute those steps in order exactly as returned; if it returns instructions, follow them with the supplied inputs. Otherwise use the result verbatim. Do not invent behavior beyond that output. On a host without code execution, treat the Parameters schema and the code below as the exact specification and never paraphrase a step. Never edit inside the generated markers; a converter-equipped host can instead restore the original file checksum-verified with the installed rapp-agent-converter/scripts/toast.py convert SKILL.md --to agent.

"""GeneratingPodcastScript -- Use this skill whenever the user asks to write, generate, or create a podcast script or podcast episode — from a topic, or from source material such as a news digest, newsletter, email review, or set of articles — and optionally convert it to audio with Azure Text-to-Speech. Handles the initial request and every follow-up refinement (source, topic, length, cast, narration) in the same task.

Generated by the rapp skill from generating-podcast-script. The RCI capsule at the bottom of this file carries the full original; `toast.py convert` restores it byte-exact."""

import json
import re
import sys

try:
    from agents.basic_agent import BasicAgent
except ImportError:  # running OUTSIDE a brainstem -- stay executable anyway.
    class BasicAgent:  # noqa: D101 - minimal stand-in, same contract
        def __init__(self, name=None, metadata=None):
            if name:
                self.name = name
            if metadata:
                self.metadata = metadata

        def perform(self, **kwargs):
            return "Not implemented."

        def system_context(self):
            return None

        def to_tool(self):
            return {"type": "function", "function": {
                "name": self.name,
                "description": self.metadata.get("description", ""),
                "parameters": self.metadata.get("parameters", {})}}

# The procedural layer, verbatim from the source capability.
INSTRUCTIONS = '# Podcast Script Generation\n\nProduces a two-host, NotebookLM-style conversational episode: a readable script\nplus a multi-voice SSML document ready for Azure Text-to-Speech.\n\n## Step 1 — Gather inputs\n\nAsk only for what is missing. Do not ask for optional fields that have safe\ndefaults.\n\n| Input | Required | Default |\n|---|---|---|\n| `source` or `topic` | Yes | — (source material such as a digest/newsletter/articles, **or** a plain topic) |\n| `duration` | No | `medium` — `short` ≈ 3 min / ~450 words, `medium` ≈ 6 min / ~900 words, `long` ≈ 12 min / ~1,800 words |\n| `language` | No | English |\n| `cast` | No | Two hosts, NOVA and MILES (see Step 4) |\n| `generate_audio` | No | Ask after the script is ready |\n\nAll word targets assume ~150 spoken words per minute. Stay within 10 percent of\ntarget.\n\nDerive a lowercase-hyphenated `<slug>` from the source or topic (max 6 words,\nASCII only).\n\n## Step 2 — Parse the source material\n\nSkip this step when the user gave a bare topic with no source material.\n\nExtract every distinct article or item. For each, capture:\n\n- headline\n- publication / source\n- date, if present\n- the core factual claim\n- any figures or quotes\n- the "so what"\n\nMerge duplicates covering the same event. Discard boilerplate, footers,\ndisclaimers, legal notices, unsubscribe text, and image captions.\n\n## Step 3 — Editorial selection\n\nRank items by newsworthiness and reader impact. Keep the top 4 to 6 for full\ntreatment (fewer for `short`, more for `long`). Group the remainder into one\nfast **rapid fire** segment. If the material has a dominant theme, lead with it\nand thread it through the episode.\n\n## Step 4 — Cast\n\nTwo hosts, always the same personalities, always distinct voices:\n\n- **NOVA** — voice `en-US-AvaMultilingualNeural`. Lead host. Warm, curious,\n  quick. Drives the agenda, asks the question the listener is thinking, reacts\n  out loud, reframes jargon into plain language. Slightly faster cadence.\n- **MILES** — voice `en-US-AndrewMultilingualNeural`. Analyst. Calm, dry,\n  precise. Supplies context, numbers, caveats, second-order implications.\n  Slightly slower, lower pitch. Occasionally pushes back on Nova.\n\nNeither host is a narrator. They talk **to each other**, not to the microphone.\n\n## Step 5 — Episode structure\n\n1. **Cold open** (15–20s) — Nova opens on the single most striking fact or\n   tension. No "welcome to the podcast", no channel branding, no music cues.\n2. **Agenda tease** (10s) — Miles lays out what they'll cover, casually.\n3. **Story segments** (60–90s each) — one item per segment. Pattern:\n   hook → the facts → why it matters → a short exchange of interpretation →\n   handoff line into the next story. Vary who leads each segment.\n4. **Rapid fire** (45s) — alternating one-liners on the leftover items, quick\n   tempo.\n5. **Close** (20–30s) — three concrete takeaways split between the two hosts,\n   then a short human sign-off. Alternate whether Nova or Miles ends.\n\n## Step 6 — Dialogue style\n\nThis is what makes it sound conversational rather than read-aloud.\n\n- Write spoken English, not written English. Contractions everywhere.\n- Keep most lines under 30 words. Break long explanations across two or three\n  turns with the other host interjecting.\n- Use real conversational connective tissue — "okay so", "right", "wait, back\n  up", "here's the part I didn't expect", "yeah, and that's the thing", "hmm".\n  Roughly one marker every 4 to 5 turns. Never let it become a tic.\n- One host regularly asks the naive clarifying question so the other can\n  explain.\n- Use one concrete analogy or comparison per complex item.\n- Genuine reactions are allowed ("that number is wild"). Invented opinions on\n  people, companies, or politics are not.\n- Never read a headline verbatim. Paraphrase it into speech.\n- Attribute clearly: "according to the Financial Times", "Reuters is\n  reporting".\n- If the source is ambiguous or a claim is unconfirmed, say so on air: "the\n  report is careful to call that unconfirmed".\n- No stage directions, no "[laughs]", no speaker labels, no markdown, no\n  emojis, no bullet points, and no URLs in the **spoken text**.\n\n## Step 7 — TTS hygiene\n\nApplies to every word that will be spoken.\n\n- Spell out anything a synthesizer would mangle: "twenty twenty-six" not\n  `2026`, "three point two billion dollars" not `$3.2B`, "about fifteen\n  percent" not `~15%`.\n- First mention of an acronym: expand it, then use the short form.\n- Letter-by-letter acronyms: `<say-as interpret-as="characters">API</say-as>`.\n- Odd proper nouns: `<sub alias="phonetic spelling">Name</sub>`.\n- A non-English name or phrase inside an English line:\n  `<lang xml:lang="fr-FR">...</lang>`.\n- Escape XML entities in all spoken text: `&` → `&amp;`, `<` → `&lt;`,\n  `>` → `&gt;`.\n- Never emit smart quotes, em dashes, asterisks, or underscores.\n\n## Step 8 — Write the readable script\n\nWrite the human-readable transcript to:\n\n```text\n/app/created/<slug>_Podcast_Script.txt\n```\n\nThis file — and only this file — may carry `NOVA:` / `MILES:` speaker labels so\na person can follow along. It contains no stage directions and no markdown. The\nspoken text itself must already satisfy Steps 6 and 7 so it can be lifted into\nSSML unchanged.\n\n## Step 9 — SSML output contract\n\nProduce the SSML document and **nothing else** — no preamble, no explanation, no\ncode fences, no trailing notes inside the artifact. Write it verbatim to:\n\n```text\n/app/created/<slug>_Podcast.ssml\n```\n\nRules:\n\n- Exactly one root `<speak>` element with `version="1.0"`,\n  `xmlns="http://www.w3.org/2001/10/synthesis"`,\n  `xmlns:mstts="http://www.w3.org/2001/mstts"`, `xml:lang="en-US"`.\n- One `<voice>` element per conversational turn. Alternate speakers. Never put\n  both hosts inside one `<voice>` element.\n- Vary delivery with `<prosody>` so it never sounds flat. Baselines: Nova\n  `rate="+6%" pitch="+2%"`, Miles `rate="-2%" pitch="-4%"`. Nudge per line to\n  match the emotion of the sentence.\n- Use `<mstts:express-as style="...">` where the voice supports it. Preferred\n  styles: `chat` for banter, `friendly` for explanation,\n  `narration-professional` for the factual core of a story, `excited` sparingly\n  for the cold open. An unsupported style is ignored by the service, so keep\n  styles optional — never structural.\n- Pauses: `<break time="250ms"/>` between turns within a segment,\n  `<break time="700ms"/>` between segments, `<break time="400ms"/>` before a\n  punchline or a pivot. Never exceed `900ms`.\n- **Every `<break>` must sit inside a `<voice>` element.** A `<break>` placed\n  between `<voice>` elements — as a direct child of `<speak>` — is invalid in a\n  multi-voice document and will fail synthesis. Two `<voice>` elements may sit\n  directly next to each other with nothing between them. To pause *between*\n  turns, put the break at the **end of the preceding turn's** text, inside that\n  turn's `<prosody>`.\n- `<emphasis level="moderate">` on at most one or two key terms per segment.\n- Keep the total document under 40,000 characters.\n\n### Shape\n\n```xml\n<speak version="1.0"\n       xmlns="http://www.w3.org/2001/10/synthesis"\n       xmlns:mstts="http://www.w3.org/2001/mstts"\n       xml:lang="en-US">\n  <voice name="en-US-AvaMultilingualNeural">\n    <mstts:express-as style="excited">\n      <prosody rate="+8%" pitch="+3%">Okay, so the number that stopped me cold\n      this morning was forty percent. <break time="300ms"/> Forty percent, in one\n      quarter. <break time="250ms"/></prosody>\n    </mstts:express-as>\n  </voice>\n  <voice name="en-US-AndrewMultilingualNeural">\n    <mstts:express-as style="chat">\n      <prosody rate="-2%" pitch="-4%">Right, and the part everyone's skipping is\n      that it's off a very small base. <break time="250ms"/> Context matters\n      here. <break time="700ms"/></prosody>\n    </mstts:express-as>\n  </voice>\n</speak>\n```\n\nNote the trailing `<break>` closing each turn: the 250ms is the gap before the\nnext turn, the 700ms is the longer gap before the next segment. Nothing sits\nbetween the two `<voice>` elements.\n\n## Step 10 — Review, then generate audio\n\nShow the user a table summarising each segment (title, one-line description,\napproximate spoken duration), print both file paths, and ask:\n\n> Would you like me to convert this to an audio file?\n\nOnly if they say yes:\n\n1. Confirm the `ConverttexttospeechwithSSML` tool is available on the agent. If\n   it is not, tell the user how to add it and stop.\n2. Call `ConverttexttospeechwithSSML` with the SSML document from Step 9 and\n   `outputFormat: riff-24khz-16bit-mono-pcm`.\n3. Decode the base64 response and save to `/app/created/<slug>_Podcast.wav`.\n\nIf the document is too large for a single call, split it at a segment boundary and synthesize each part separately.\nWhen stitching, do NOT concatenate decoded RIFF/WAV bytes; instead, append the audio at the PCM-frame level (e.g., via Python’s `wave` module) and write a single valid `/app/created/<slug>_Podcast.wav` with one header.\n\n```python\nimport base64\n\nwith open('<tool_output_file>', 'r') as f:\n    content = f.read().strip()\n\naudio_bytes = base64.b64decode(content)\n\nwith open('/app/created/<slug>_Podcast.wav', 'wb') as f:\n    f.write(audio_bytes)\n```\n\n## Step 11 — Final report\n\nAlways end with:\n\n| Item | Details |\n|---|---|\n| Script file | `/app/created/<slug>_Podcast_Script.txt` |\n| SSML file | `/app/created/<slug>_Podcast.ssml` |\n| Word count / estimated duration | actual vs. target at ~150 wpm |\n| Items covered | full segments + rapid-fire count |\n| Voices | `en-US-AvaMultilingualNeural` (Nova), `en-US-AndrewMultilingualNeural` (Miles) |\n| Audio file | `/app/created/<slug>_Podcast.wav` *(only if audio generated)* |'

# Ordered commands lifted verbatim from the capability's own documentation.
STEPS = []


class GeneratingPodcastScriptAgent(BasicAgent):
    def __init__(self):
        self.name = 'GeneratingPodcastScript'
        self.metadata = {
          "name": "GeneratingPodcastScript",
          "description": "Use this skill whenever the user asks to write, generate, or create a podcast script or podcast episode \u2014 from a topic, or from source material such as a news digest, newsletter, email review, or set of articles \u2014 and optionally convert it to audio with Azure Text-to-Speech. Handles the initial request and every follow-up refinement (source, topic, length, cast, narration) in the same task.",
          "parameters": {
            "type": "object",
            "properties": {},
            "required": []
          }
        }
        super().__init__(name=self.name, metadata=self.metadata)

    def perform(self, **kwargs):  # toaster:generated-perform
        return json.dumps({"status": "ok", "instructions": INSTRUCTIONS,
                           "inputs": kwargs,
                           "note": "Prose-only capability: follow INSTRUCTIONS "
                                   "with the given inputs."}, indent=2)

if __name__ == "__main__":
    #     echo '{"arg": "value"}' | python3 generating_podcast_script_agent.py
    #     python3 generating_podcast_script_agent.py '{"arg": "value"}'
    #     python3 generating_podcast_script_agent.py --tool          # emit the JSON tool contract
    _a = sys.argv[1:]
    if _a and _a[0] == "--tool":
        print(json.dumps(GeneratingPodcastScriptAgent().to_tool(), indent=2))
    else:
        _raw = _a[0] if _a else (sys.stdin.read().strip() or "{}")
        print(GeneratingPodcastScriptAgent().perform(**json.loads(_raw)))

# rci-capsule:v1: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