Plunder & Riffle Meme Engine is an MCP server published by io.github.davenamovich, hosted at bot.plunderandriffle.com. When last checked on 2026-11-22 it was working: it answers and lists its tools. It declares 22 tools, including generate_ad_creative, request_mp4, list_motion_styles, create_funnel, create_ridiculous.
Memes, ads and Jobless Joe shows from one prompt: a preview page first, an MP4 after approval.
4 connections through 4 shared values. Tap any node.
Inferred from the host name and headers: what a server runs on, not who runs it.
One ad creative in one language, in the `format` you pass (omitted = single meme). Returns a TEMP PAGE preview (not an MP4), PNG URLs (adAssets, media_url), post_copy and a referralUrl. Costs the format's list credits; BYOK: $0.01. Optional `motion_style` (a banked style id from list_motion_styles) leads a moving format's motion design with that style's full director prompt — same price.
| Parameter | Type | Description |
|---|---|---|
| ref | string | Your referral code (returned in referralUrl). |
| brand | string | Brand key (default house brand). |
| brief * | string | What the ad is about: product, hook, angle. |
| offer | string | Offer or call to action to work into the creative. |
None yet. Changes appear after a server's second daily check.
Squares are this server's values; dots are other servers sharing them. A shared address or publisher is a lead, not proof of a common operator.
| sizes | array | PNG sizes (default all). Not the creative `format`. |
| format | string | Creative format — set it to what was asked; omitted = single meme. iMessage / text thread → "hybrid"; carousel / slideshow → "carousel"; hot take → "hottake"; confession → "confession"; single meme / image → "image". |
| audience | string |
| language | string | Default English. |
| allFrames | boolean | Export every slide (max 10). Default true for carousel/hybrid. |
| brandProfile | object | Ad-hoc brand voice, overrides `brand`: name, voice (≤1000), hashtag, token, ctaStyle {tone, action (no URL), linkPlacement first-comment|caption, avoid[], examples[]}. |
| motion_style | string | Optional banked motion-design style id (see list_motion_styles). For the moving formats (carousel, hybrid) the style's full director prompt leads the generation instead of the automatic style pick-list. Same price as the call without it. Ignored by still formats. |
| includeAssets | boolean | Render PNGs and return URLs (default true; false for single memes). |
{
"name": "generate_ad_creative",
"title": "Make a meme or ad",
"description": "One ad creative in one language, in the `format` you pass (omitted = single meme). Returns a TEMP PAGE preview (not an MP4), PNG URLs (adAssets, media_url), post_copy and a referralUrl. Costs the format's list credits; BYOK: $0.01. Optional `motion_style` (a banked style id from list_motion_styles) leads a moving format's motion design with that style's full director prompt — same price.",
"inputSchema": {
"type": "object",
"required": [
"brief"
],
"properties": {
"ref": {
"type": "string",
"description": "Your referral code (returned in referralUrl)."
},
"brand": {
"type": "string",
"description": "Brand key (default house brand)."
},
"brief": {
"type": "string",
"description": "What the ad is about: product, hook, angle."
},
"offer": {
"type": "string",
"description": "Offer or call to action to work into the creative."
},
"sizes": {
"type": "array",
"items": {
"enum": [
"square",
"feed",
"story"
],
"type": "string"
},
"description": "PNG sizes (default all). Not the creative `format`."
},
"format": {
"enum": [
"image",
"carousel",
"confession",
"hottake",
"hybrid",
"xaiver-reply"
],
"type": "string",
"description": "Creative format — set it to what was asked; omitted = single meme. iMessage / text thread → \"hybrid\"; carousel / slideshow → \"carousel\"; hot take → \"hottake\"; confession → \"confession\"; single meme / image → \"image\"."
},
"audience": {
"type": "string"
},
"language": {
"type": "string",
"description": "Default English."
},
"allFrames": {
"type": "boolean",
"description": "Export every slide (max 10). Default true for carousel/hybrid."
},
"brandProfile": {
"type": "object",
"required": [
"name"
],
"properties": {
"name": {
"type": "string"
}
},
"description": "Ad-hoc brand voice, overrides `brand`: name, voice (≤1000), hashtag, token, ctaStyle {tone, action (no URL), linkPlacement first-comment|caption, avoid[], examples[]}."
},
"motion_style": {
"enum": [
"flat-vector",
"line-art",
"isometric",
"cel-boil",
"collage",
"liquid-motion",
"shape-morph",
"bauhaus",
"variety-captions",
"sticker-explainer"
],
"type": "string",
"description": "Optional banked motion-design style id (see list_motion_styles). For the moving formats (carousel, hybrid) the style's full director prompt leads the generation instead of the automatic style pick-list. Same price as the call without it. Ignored by still formats."
},
"includeAssets": {
"type": "boolean",
"description": "Render PNGs and return URLs (default true; false for single memes)."
}
}
},
"annotations": {
"destructiveHint": false
}
}AFTER a human has approved a preview, start rendering an MP4 of that temp page (25 credits = $0.25; BYOK: $0.01 with your own Railway worker saved or when full BYOK — every required key saved at /vip-setup — else the full $0.25; refunded on failure). Takes 1–3 min. To skip polling pass callback_url and/or wait_seconds; otherwise poll get_job_status (kind mp4) with the jobId. Never render MP4s for unapproved previews.
| Parameter | Type | Description |
|---|---|---|
| intro | string | Optional: "joe_call" opens the video with Joe calling on a phone to set up your video. |
| pageUrl * | string | The approved pageUrl from generate_ad_creative. |
| outro_url | ["string","null"] | Optional: an https MP4 (up to 60 s) to play after this video instead of your account outro. null = no outro for this render. |
| orientation | string | Default vertical 9:16. Use "landscape" (16:9, no intro/outro clips) for a 16:9 page. |
| callback_url | string | Optional https webhook. Receives the signed done/failed result; the response carries the webhook secret once. |
| wait_seconds | number | Optional, 0-100. Wait this long for the MP4 and return it in the same call. |
{
"name": "request_mp4",
"title": "Render an approved preview as MP4",
"description": "AFTER a human has approved a preview, start rendering an MP4 of that temp page (25 credits = $0.25; BYOK: $0.01 with your own Railway worker saved or when full BYOK — every required key saved at /vip-setup — else the full $0.25; refunded on failure). Takes 1–3 min. To skip polling pass callback_url and/or wait_seconds; otherwise poll get_job_status (kind mp4) with the jobId. Never render MP4s for unapproved previews.",
"inputSchema": {
"type": "object",
"required": [
"pageUrl"
],
"properties": {
"intro": {
"type": "string",
"description": "Optional: \"joe_call\" opens the video with Joe calling on a phone to set up your video."
},
"pageUrl": {
"type": "string",
"description": "The approved pageUrl from generate_ad_creative."
},
"outro_url": {
"type": [
"string",
"null"
],
"description": "Optional: an https MP4 (up to 60 s) to play after this video instead of your account outro. null = no outro for this render."
},
"orientation": {
"enum": [
"vertical",
"landscape"
],
"type": "string",
"description": "Default vertical 9:16. Use \"landscape\" (16:9, no intro/outro clips) for a 16:9 page."
},
"callback_url": {
"type": "string",
"description": "Optional https webhook. Receives the signed done/failed result; the response carries the webhook secret once."
},
"wait_seconds": {
"type": "number",
"description": "Optional, 0-100. Wait this long for the MP4 and return it in the same call."
}
}
},
"annotations": {
"destructiveHint": false
}
}Free discovery for the engine's banked motion craft (ported 2026-11-20): motion-design style presets (mg-styles-15), scene grammars for narrated pieces (huashu-art-motion) and measured easings (motionmaxxing). These banks ride into slideshow / hybrid generation automatically; call this to see them, or pass `style` (a style id like "flat-vector") to get that style's full director prompt for building a piece in it yourself. Pass `section`: "grammars" or "eases" for those banks in full. No key needed, nothing charged.
| Parameter | Type | Description |
|---|---|---|
| style | string | A motion style id (from the styles list). Returns the style preset plus its full director prompt. |
| section | string | "styles" (default), "grammars" or "eases". |
{
"name": "list_motion_styles",
"title": "List motion styles",
"description": "Free discovery for the engine's banked motion craft (ported 2026-11-20): motion-design style presets (mg-styles-15), scene grammars for narrated pieces (huashu-art-motion) and measured easings (motionmaxxing). These banks ride into slideshow / hybrid generation automatically; call this to see them, or pass `style` (a style id like \"flat-vector\") to get that style's full director prompt for building a piece in it yourself. Pass `section`: \"grammars\" or \"eases\" for those banks in full. No key needed, nothing charged.",
"inputSchema": {
"type": "object",
"properties": {
"style": {
"type": "string",
"description": "A motion style id (from the styles list). Returns the style preset plus its full director prompt."
},
"section": {
"type": "string",
"description": "\"styles\" (default), \"grammars\" or \"eases\"."
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false
}
}/funnel: ONE offer link becomes voiced content whose CTA is that link: captions (+ an iMessage thread; voiced typography page, no memes), memes (timed-caption voiced page with memes, an iMessage thread, + a picture or iMessage meme) or a stick-figure story; format bestof builds a silent day of content (2 memes, an iMessage, a slideshow, a ranked list) AND publishes one share page bundling everything (dayPack.shareUrl); format share rebuilds that page. Reads the page (never invents; unreadable = status needs-description, then call with funnel_id + description). Omit format to analyze first; switch with funnel_id + format, no new link. 10 credits to read; captions 25, memes 35, stickfigures 25, bestof 60 (BYOK $0.01/call); refunded on failure. Poll get_job_status (kind funnel, jobId = funnelId). Previews only: show offer.url under them; no posting or MP4 without the human's OK. For an MP4, pass mp4PageUrl (the creative — a voiced video, the show, a meme) to request_mp4, never dayPack.shareUrl, which is the carousel bundle.
| Parameter | Type | Description |
|---|---|---|
| cta | string | Button label under the preview. Default: from the product name. |
| url | string | The offer link, kept exactly (referral params untouched). Not needed with funnel_id. |
| kind | string | memes only: meme | imessage (default: alternates). |
| tone | string | |
| angle | string | |
| format | string | captions | memes | stickfigures | all | bestof |
| script | string | Your own words, spoken as is. |
| context | string | |
| rebuild | boolean | With funnel_id: the same funnel again, made fresh. |
| audience | string | |
| language | string | |
| platform | string | |
| funnel_id | string | Earlier funnelId: switch format, regenerate, describe or rebuild without the link. |
| objection | string | |
| variation | string | same = regenerate, next = new angle (default same). |
| description | string | What the offer is and who it's for, when the page can't be read. |
{
"name": "create_funnel",
"title": "Content from one offer link",
"description": "/funnel: ONE offer link becomes voiced content whose CTA is that link: captions (+ an iMessage thread; voiced typography page, no memes), memes (timed-caption voiced page with memes, an iMessage thread, + a picture or iMessage meme) or a stick-figure story; format bestof builds a silent day of content (2 memes, an iMessage, a slideshow, a ranked list) AND publishes one share page bundling everything (dayPack.shareUrl); format share rebuilds that page. Reads the page (never invents; unreadable = status needs-description, then call with funnel_id + description). Omit format to analyze first; switch with funnel_id + format, no new link. 10 credits to read; captions 25, memes 35, stickfigures 25, bestof 60 (BYOK $0.01/call); refunded on failure. Poll get_job_status (kind funnel, jobId = funnelId). Previews only: show offer.url under them; no posting or MP4 without the human's OK. For an MP4, pass mp4PageUrl (the creative — a voiced video, the show, a meme) to request_mp4, never dayPack.shareUrl, which is the carousel bundle.",
"inputSchema": {
"type": "object",
"properties": {
"cta": {
"type": "string",
"description": "Button label under the preview. Default: from the product name."
},
"url": {
"type": "string",
"description": "The offer link, kept exactly (referral params untouched). Not needed with funnel_id."
},
"kind": {
"type": "string",
"description": "memes only: meme | imessage (default: alternates)."
},
"tone": {
"type": "string"
},
"angle": {
"type": "string"
},
"format": {
"type": "string",
"description": "captions | memes | stickfigures | all | bestof"
},
"script": {
"type": "string",
"description": "Your own words, spoken as is."
},
"context": {
"type": "string"
},
"rebuild": {
"type": "boolean",
"description": "With funnel_id: the same funnel again, made fresh."
},
"audience": {
"type": "string"
},
"language": {
"type": "string"
},
"platform": {
"type": "string"
},
"funnel_id": {
"type": "string",
"description": "Earlier funnelId: switch format, regenerate, describe or rebuild without the link."
},
"objection": {
"type": "string"
},
"variation": {
"type": "string",
"description": "same = regenerate, next = new angle (default same)."
},
"description": {
"type": "string",
"description": "What the offer is and who it's for, when the page can't be read."
}
}
},
"annotations": {
"destructiveHint": false
}
}The /ridiculous command: one ~30s episode of Jobless Joe's meme reaction show — Joe roasts the topic in five scenes while Ralph acts it out on a giant TV. `topic` is a subject, a link (roasted as a website autopsy), a business, or "roast my competitor's ad". Optional `format` forces a segment. 25 credits ($0.25; BYOK: $0.01), refunded on failure. Async: poll get_job_status (kind sitdown), 1–2 min. Done carries pageUrl (a preview, no MP4): show it; request_mp4 only after the human approves. Done also carries `format` and `altEnding`.
| Parameter | Type | Description |
|---|---|---|
| topic * | string | What the episode roasts (≤400 chars): a topic, a link, a business, or e.g. "roast my competitor's ad". |
| format | string | Force a segment. Default: picked from the topic (a link is a website autopsy). |
| language | string | Default English. |
| alt_ending | boolean | Use the episode's alternate last line instead of the main one. |
{
"name": "create_ridiculous",
"title": "Jobless Joe reaction show",
"description": "The /ridiculous command: one ~30s episode of Jobless Joe's meme reaction show — Joe roasts the topic in five scenes while Ralph acts it out on a giant TV. `topic` is a subject, a link (roasted as a website autopsy), a business, or \"roast my competitor's ad\". Optional `format` forces a segment. 25 credits ($0.25; BYOK: $0.01), refunded on failure. Async: poll get_job_status (kind sitdown), 1–2 min. Done carries pageUrl (a preview, no MP4): show it; request_mp4 only after the human approves. Done also carries `format` and `altEnding`.",
"inputSchema": {
"type": "object",
"required": [
"topic"
],
"properties": {
"topic": {
"type": "string",
"description": "What the episode roasts (≤400 chars): a topic, a link, a business, or e.g. \"roast my competitor's ad\"."
},
"format": {
"enum": [
"ai",
"adcrimes",
"jobless",
"autopsy",
"court",
"plug"
],
"type": "string",
"description": "Force a segment. Default: picked from the topic (a link is a website autopsy)."
},
"language": {
"type": "string",
"description": "Default English."
},
"alt_ending": {
"type": "boolean",
"description": "Use the episode's alternate last line instead of the main one."
}
}
},
"annotations": {
"destructiveHint": false
}
}Telegram /present. 16:9 Joe presentation. ONE of topic, script, or audio_url. Poll get_job_status (sitdown); request_mp4 orientation "landscape". 25 credits ($0.25; BYOK $0.01).
| Parameter | Type | Description |
|---|---|---|
| game | string | awareness (default) or conversion |
| topic | string | |
| length | string | topic only |
| script | string | |
| template | string | one of ["documentary","investor","educational","marketing","productExplainer"] |
| audio_url | string | https mp3/wav/m4a/mp4, ≤24 MB |
{
"name": "create_present",
"title": "Presentation video",
"description": "Telegram /present. 16:9 Joe presentation. ONE of topic, script, or audio_url. Poll get_job_status (sitdown); request_mp4 orientation \"landscape\". 25 credits ($0.25; BYOK $0.01).",
"inputSchema": {
"type": "object",
"properties": {
"game": {
"enum": [
"awareness",
"conversion"
],
"type": "string",
"description": "awareness (default) or conversion"
},
"topic": {
"type": "string"
},
"length": {
"enum": [
"short",
"long"
],
"type": "string",
"description": "topic only"
},
"script": {
"type": "string"
},
"template": {
"enum": [
"documentary",
"investor",
"educational",
"marketing",
"productExplainer"
],
"type": "string"
},
"audio_url": {
"type": "string",
"description": "https mp3/wav/m4a/mp4, ≤24 MB"
}
}
},
"annotations": {
"destructiveHint": false
}
}Joe at the whiteboard: a narrated pitch/results page built from YOUR deck `config` (title, presenter {character, voiceRef}, business, scenes; chart scenes bignum/bars/versus draw the config's own numbers as SVG — the engine invents nothing). `voice_id` locks the narration voice for the call (overrides the config's voiceRef); scenes are auto-tagged chart/sketch/cta and the result reports the voice used + the tags. Returns pageUrl (preview); request_mp4 only after the human approves. 25 credits ($0.25; BYOK $0.01), refunded on failure.
| Parameter | Type | Description |
|---|---|---|
| config * | object | The full whiteboard-pitch deck config. Needs a presenter with a voiceRef and at least one chart scene with valid data. |
| voice_id | string | Narration voice lock for this call — overrides config.presenter.voiceRef. |
{
"name": "create_whiteboard_pitch",
"title": "Whiteboard pitch video",
"description": "Joe at the whiteboard: a narrated pitch/results page built from YOUR deck `config` (title, presenter {character, voiceRef}, business, scenes; chart scenes bignum/bars/versus draw the config's own numbers as SVG — the engine invents nothing). `voice_id` locks the narration voice for the call (overrides the config's voiceRef); scenes are auto-tagged chart/sketch/cta and the result reports the voice used + the tags. Returns pageUrl (preview); request_mp4 only after the human approves. 25 credits ($0.25; BYOK $0.01), refunded on failure.",
"inputSchema": {
"type": "object",
"required": [
"config"
],
"properties": {
"config": {
"type": "object",
"description": "The full whiteboard-pitch deck config. Needs a presenter with a voiceRef and at least one chart scene with valid data."
},
"voice_id": {
"type": "string",
"description": "Narration voice lock for this call — overrides config.presenter.voiceRef."
}
}
},
"annotations": {
"destructiveHint": false
}
}Telegram /joenews (not /news). Live sources. Desk/tv, 45–120s. mode or topic; short|standard|deep; language. Refunds if none answer. 25 credits ($0.25; BYOK $0.01). Poll get_job_status (sitdown): pageUrl, caption, sources. One outlet is reported. MP4 after approval.
| Parameter | Type | Description |
|---|---|---|
| mode | string | Default general (top stories). Ignored with a topic. |
| pages | array | ≤3 https pages that play on Joe's TV: temp /m/ page, .mp4/.gif/.png, here.now or plunderandriffle.com page. |
| topic | string | Report on this instead (≤200 chars). |
| layout | string | desk (default): Joe + Phony showing each story's picture; tv: the giant TV. |
| length | string | ≈45s · ≈90s (default) · ≈2 min. |
| language | string | Default English. |
| audio_url | string | https mp3/wav/m4a that plays first; Joe reacts. |
{
"name": "create_news",
"title": "Jobless Joe News",
"description": "Telegram /joenews (not /news). Live sources. Desk/tv, 45–120s. mode or topic; short|standard|deep; language. Refunds if none answer. 25 credits ($0.25; BYOK $0.01). Poll get_job_status (sitdown): pageUrl, caption, sources. One outlet is reported. MP4 after approval.",
"inputSchema": {
"type": "object",
"properties": {
"mode": {
"enum": [
"general",
"ai",
"crypto",
"business",
"internet",
"conspiracy"
],
"type": "string",
"description": "Default general (top stories). Ignored with a topic."
},
"pages": {
"type": "array",
"items": {
"type": "string"
},
"description": "≤3 https pages that play on Joe's TV: temp /m/ page, .mp4/.gif/.png, here.now or plunderandriffle.com page."
},
"topic": {
"type": "string",
"description": "Report on this instead (≤200 chars)."
},
"layout": {
"enum": [
"desk",
"tv"
],
"type": "string",
"description": "desk (default): Joe + Phony showing each story's picture; tv: the giant TV."
},
"length": {
"enum": [
"short",
"standard",
"deep"
],
"type": "string",
"description": "≈45s · ≈90s (default) · ≈2 min."
},
"language": {
"type": "string",
"description": "Default English."
},
"audio_url": {
"type": "string",
"description": "https mp3/wav/m4a that plays first; Joe reacts."
}
}
},
"annotations": {
"destructiveHint": false
}
}The /sitdown bestof command: 1–5 temp pages or direct media links play, silent, on the sidekick's screen while the host counts them down — one voiced best-of reel. Give `links` and an optional `theme`. 25 credits ($0.25; BYOK: $0.01), refunded on failure. Async: poll get_job_status (kind sitdown), 1–2 min. Done carries pageUrl (a preview, no MP4): show it; request_mp4 only after the human approves.
| Parameter | Type | Description |
|---|---|---|
| links * | array | 1–5 https links: a temp page (/m/<id>) or a direct .mp4/.gif/.png. A whitespace-separated string works too. |
| scene | string | The set (default: the owner's saved one). |
| theme | string | Optional theme the countdown hangs on, e.g. "job hunt fails" (≤120 chars). |
| outfit | string | Default: the brand's look. |
| language | string | Default English. |
| accessory | string | 2–5 words on the host's shirt (≤40 chars). Implies outfit: tee. |
{
"name": "create_best_of",
"title": "Best-of reel from links",
"description": "The /sitdown bestof command: 1–5 temp pages or direct media links play, silent, on the sidekick's screen while the host counts them down — one voiced best-of reel. Give `links` and an optional `theme`. 25 credits ($0.25; BYOK: $0.01), refunded on failure. Async: poll get_job_status (kind sitdown), 1–2 min. Done carries pageUrl (a preview, no MP4): show it; request_mp4 only after the human approves.",
"inputSchema": {
"type": "object",
"required": [
"links"
],
"properties": {
"links": {
"type": "array",
"items": {
"type": "string"
},
"description": "1–5 https links: a temp page (/m/<id>) or a direct .mp4/.gif/.png. A whitespace-separated string works too."
},
"scene": {
"enum": [
"desk",
"lounge",
"podcast",
"news",
"stage"
],
"type": "string",
"description": "The set (default: the owner's saved one)."
},
"theme": {
"type": "string",
"description": "Optional theme the countdown hangs on, e.g. \"job hunt fails\" (≤120 chars)."
},
"outfit": {
"enum": [
"tie",
"tee"
],
"type": "string",
"description": "Default: the brand's look."
},
"language": {
"type": "string",
"description": "Default English."
},
"accessory": {
"type": "string",
"description": "2–5 words on the host's shirt (≤40 chars). Implies outfit: tee."
}
}
},
"annotations": {
"destructiveHint": false
}
}The /shop push pack: any product link or description → researched, ready-to-post copy (hooks, meme lines, iMessage thread, CTA, hashtags, first comment) plus up to 3 creatives (TEMP PAGE previews + PNGs + post_copy). 10 credits for the context + each format's list credits; BYOK: $0.01. Each part refunded on its own failure. ~30–90s.
| Parameter | Type | Description |
|---|---|---|
| ref | string | Your referral code (returned in referralUrl). |
| brand | string | Brand key (default house brand). |
| input * | string | A product/page link, or a description of the product when there is no link (≤1500 chars). |
| sizes | array | PNG sizes (default all). Not the creative `format`. |
| language | string | Default English. |
| allFrames | boolean | Export every slide (max 10). Default true for carousel/hybrid. |
| creatives | array | Creative formats to build (max 3); omitted = one single meme, [] = copy only. iMessage / text thread → "hybrid"; carousel / slideshow → "carousel"; hot take → "hottake"; confession → "confession"; single meme / image → "image". |
| direction | string | Optional angle/audience/tone, e.g. "for burnt-out PMs". |
| brandProfile | object | Ad-hoc brand voice, overrides `brand`: name, voice (≤1000), hashtag, token, ctaStyle {tone, action (no URL), linkPlacement first-comment|caption, avoid[], examples[]}. |
| includeAssets | boolean | Render PNGs and return URLs (default true; false for single memes). |
{
"name": "shop",
"title": "Product ad copy and creatives",
"description": "The /shop push pack: any product link or description → researched, ready-to-post copy (hooks, meme lines, iMessage thread, CTA, hashtags, first comment) plus up to 3 creatives (TEMP PAGE previews + PNGs + post_copy). 10 credits for the context + each format's list credits; BYOK: $0.01. Each part refunded on its own failure. ~30–90s.",
"inputSchema": {
"type": "object",
"required": [
"input"
],
"properties": {
"ref": {
"type": "string",
"description": "Your referral code (returned in referralUrl)."
},
"brand": {
"type": "string",
"description": "Brand key (default house brand)."
},
"input": {
"type": "string",
"description": "A product/page link, or a description of the product when there is no link (≤1500 chars)."
},
"sizes": {
"type": "array",
"items": {
"enum": [
"square",
"feed",
"story"
],
"type": "string"
},
"description": "PNG sizes (default all). Not the creative `format`."
},
"language": {
"type": "string",
"description": "Default English."
},
"allFrames": {
"type": "boolean",
"description": "Export every slide (max 10). Default true for carousel/hybrid."
},
"creatives": {
"type": "array",
"items": {
"enum": [
"image",
"carousel",
"confession",
"hottake",
"hybrid",
"xaiver-reply"
],
"type": "string"
},
"description": "Creative formats to build (max 3); omitted = one single meme, [] = copy only. iMessage / text thread → \"hybrid\"; carousel / slideshow → \"carousel\"; hot take → \"hottake\"; confession → \"confession\"; single meme / image → \"image\"."
},
"direction": {
"type": "string",
"description": "Optional angle/audience/tone, e.g. \"for burnt-out PMs\"."
},
"brandProfile": {
"type": "object",
"required": [
"name"
],
"properties": {
"name": {
"type": "string"
}
},
"description": "Ad-hoc brand voice, overrides `brand`: name, voice (≤1000), hashtag, token, ctaStyle {tone, action (no URL), linkPlacement first-comment|caption, avoid[], examples[]}."
},
"includeAssets": {
"type": "boolean",
"description": "Render PNGs and return URLs (default true; false for single memes)."
}
}
},
"annotations": {
"destructiveHint": false
}
}Telegram /before for ONE prospect (site, x.com/linkedin link or a sentence; never a data program's own site): Joe's welcome + five custom examples, in their site's language. campaign_id reuses a /2step's research. product pins the iMessage to a specific product/service. competitor frames the iMessage as a contrast against a named rival. deep:true crawls subpages + runs competitor-intel search before writing. Async: poll get_job_status kind demo_pack. job_id = rebuild from the same input. 50 credits ($0.50; BYOK $0.01). song_url adds a bonus song video (+30 credits). script replaces Joe's welcome script; audio_url plays a recording instead of TTS. feed:true or offer = the old text pitch: pageUrl, a 24h shareUrl with no link back, and an id for create_followup. 10 credits. feed/offer is async too: poll get_job_status kind before_text every 15-20 s.
| Parameter | Type | Description |
|---|---|---|
| deep | boolean | Deep research: crawl subpages + search competitors. |
| feed | boolean | |
| angle | string | The one argument Joe makes in the video (default auto: picked from the research). |
| input | string | Website, x.com / linkedin link, or a sentence about the person (max 1500 chars). |
| offer | string | What the sender pitches (implies feed). |
| job_id | string | |
| script | string | Your script for the welcome video (800 characters max). Not with feed/offer. |
| cta_url | string | https link the button opens, with cta_label. Default: the /options page, keeping the sender's ref. A Bankr link is not used; a t.me link becomes telegram_bot. |
| product | string | Product or service to feature in the iMessage. |
| language | string | Write and speak the pitch in this language (default: their site's). |
| song_url | string | https song link: bonus song video, +30 credits, shown as a sixth phone. Not with feed/offer. |
| audio_url | string | https mp3/wav/m4a played instead of TTS. Not with feed/offer. |
| cta_label | string | Button words (40 characters max), with cta_url. Default: SEE YOUR OPTIONS →. |
| competitor | string | Competitor or rival to contrast against in the iMessage. |
| campaign_id | string | |
| video_style | string | one of ["joe","stickman","memes","captions"] |
| telegram_bot | string | Their Telegram bot (@name or t.me link): its own button, "Chat with @name on Telegram", under the main one. |
{
"name": "create_before_pitch",
"title": "Before pitch for one business",
"description": "Telegram /before for ONE prospect (site, x.com/linkedin link or a sentence; never a data program's own site): Joe's welcome + five custom examples, in their site's language. campaign_id reuses a /2step's research. product pins the iMessage to a specific product/service. competitor frames the iMessage as a contrast against a named rival. deep:true crawls subpages + runs competitor-intel search before writing. Async: poll get_job_status kind demo_pack. job_id = rebuild from the same input. 50 credits ($0.50; BYOK $0.01). song_url adds a bonus song video (+30 credits). script replaces Joe's welcome script; audio_url plays a recording instead of TTS. feed:true or offer = the old text pitch: pageUrl, a 24h shareUrl with no link back, and an id for create_followup. 10 credits. feed/offer is async too: poll get_job_status kind before_text every 15-20 s.",
"inputSchema": {
"type": "object",
"properties": {
"deep": {
"type": "boolean",
"description": "Deep research: crawl subpages + search competitors."
},
"feed": {
"type": "boolean"
},
"angle": {
"enum": [
"auto",
"fake-ugc",
"regulars",
"trust",
"expertise",
"time",
"noticed",
"creator",
"community",
"b2b",
"health"
],
"type": "string",
"description": "The one argument Joe makes in the video (default auto: picked from the research)."
},
"input": {
"type": "string",
"description": "Website, x.com / linkedin link, or a sentence about the person (max 1500 chars)."
},
"offer": {
"type": "string",
"description": "What the sender pitches (implies feed)."
},
"job_id": {
"type": "string"
},
"script": {
"type": "string",
"description": "Your script for the welcome video (800 characters max). Not with feed/offer."
},
"cta_url": {
"type": "string",
"description": "https link the button opens, with cta_label. Default: the /options page, keeping the sender's ref. A Bankr link is not used; a t.me link becomes telegram_bot."
},
"product": {
"type": "string",
"description": "Product or service to feature in the iMessage."
},
"language": {
"type": "string",
"description": "Write and speak the pitch in this language (default: their site's)."
},
"song_url": {
"type": "string",
"description": "https song link: bonus song video, +30 credits, shown as a sixth phone. Not with feed/offer."
},
"audio_url": {
"type": "string",
"description": "https mp3/wav/m4a played instead of TTS. Not with feed/offer."
},
"cta_label": {
"type": "string",
"description": "Button words (40 characters max), with cta_url. Default: SEE YOUR OPTIONS →."
},
"competitor": {
"type": "string",
"description": "Competitor or rival to contrast against in the iMessage."
},
"campaign_id": {
"type": "string"
},
"video_style": {
"enum": [
"joe",
"stickman",
"memes",
"captions"
],
"type": "string"
},
"telegram_bot": {
"type": "string",
"description": "Their Telegram bot (@name or t.me link): its own button, \"Chat with @name on Telegram\", under the main one."
}
}
},
"annotations": {
"destructiveHint": false
}
}The five examples of a FINISHED create_before_pitch as a ~20 second vertical scroll video preview, with a minor track of Joe reacting like a friend going through them (no new examples are generated). look: feed (a For You scroll), phones (a row of phones) or chat (a group chat sending them around). Async: poll get_job_status kind scroll (seconds muted, about a minute with the track). Done carries pageUrl and joeSays: show the human; only after they approve call request_mp4 with the pageUrl. No credits for the preview.
| Parameter | Type | Description |
|---|---|---|
| look | string | Default: picked from the brand's seed. |
| mute | boolean | No Joe track. |
| seed | integer | Another cut of the same five (same look + seed = same cut). |
| job_id * | string | The jobId of a done create_before_pitch (pack_…). |
{
"name": "create_before_scroll",
"title": "Scroll video of a pitch's examples",
"description": "The five examples of a FINISHED create_before_pitch as a ~20 second vertical scroll video preview, with a minor track of Joe reacting like a friend going through them (no new examples are generated). look: feed (a For You scroll), phones (a row of phones) or chat (a group chat sending them around). Async: poll get_job_status kind scroll (seconds muted, about a minute with the track). Done carries pageUrl and joeSays: show the human; only after they approve call request_mp4 with the pageUrl. No credits for the preview.",
"inputSchema": {
"type": "object",
"required": [
"job_id"
],
"properties": {
"look": {
"enum": [
"feed",
"phones",
"chat"
],
"type": "string",
"description": "Default: picked from the brand's seed."
},
"mute": {
"type": "boolean",
"description": "No Joe track."
},
"seed": {
"type": "integer",
"minimum": 0,
"description": "Another cut of the same five (same look + seed = same cut)."
},
"job_id": {
"type": "string",
"description": "The jobId of a done create_before_pitch (pack_…)."
}
}
},
"annotations": {
"destructiveHint": false
}
}The memes of a FINISHED create_before_pitch as a ~30 second 1980s retro infomercial preview: each meme on an old TV, a billboard or a drive-in screen, then in close-up with retro actors (nothing new is generated). Async: poll get_job_status kind infomercial. Done carries pageUrl: show the human; only after they approve call request_mp4 with it. No credits for the preview.
| Parameter | Type | Description |
|---|---|---|
| mute | boolean | No voices. |
| seed | integer | Another take: other memes and sets. |
| job_id * | string | The jobId of a done create_before_pitch (pack_…). |
{
"name": "create_infomercial",
"title": "Retro infomercial of a pitch's memes",
"description": "The memes of a FINISHED create_before_pitch as a ~30 second 1980s retro infomercial preview: each meme on an old TV, a billboard or a drive-in screen, then in close-up with retro actors (nothing new is generated). Async: poll get_job_status kind infomercial. Done carries pageUrl: show the human; only after they approve call request_mp4 with it. No credits for the preview.",
"inputSchema": {
"type": "object",
"required": [
"job_id"
],
"properties": {
"mute": {
"type": "boolean",
"description": "No voices."
},
"seed": {
"type": "integer",
"minimum": 0,
"description": "Another take: other memes and sets."
},
"job_id": {
"type": "string",
"description": "The jobId of a done create_before_pitch (pack_…)."
}
}
},
"annotations": {
"destructiveHint": false
}
}The follow-up buttons over MCP: rebuild what a finished video ALREADY said as another post — same subject, same points, nothing new invented. Pass ONE source: `job_id` (a done sitdown / phone / best-of job), `source_id` (the `id` a create_before_pitch returned), or `context` (the text to build on). `format` is the output: meme, slideshow, hottake (10 credits, $0.10; BYOK: $0.01) or imessage (20 credits, $0.20). The email follow-up is the two-step outreach note (followUpGate), not this page. Returns the new pageUrl directly (no polling); show it, and only after the human approves call request_mp4 with it. Refunded on failure. Nothing is sent.
| Parameter | Type | Description |
|---|---|---|
| kind | string | With context: what the material came from (default note). |
| title | string | With context: what it is, e.g. "Pitch for Maria". |
| format * | string | meme = one image; imessage = a narrated thread. |
| job_id | string | A FINISHED sitdown or phone jobId — its own script becomes the material. |
| context | string | Material to build from when there is no id: the script, the pitch text or the key points (≤4000 chars). |
| language | string | With context: output language. A job/pitch keeps its own. |
| source_id | string | The `id` from create_before_pitch — the pitch is remembered for 2h. |
| instruction | string | Optional one-line direction for THIS piece (e.g. "make it about the interview, not the layoff"), ≤300 chars. |
{
"name": "create_followup",
"title": "Another post from a finished job",
"description": "The follow-up buttons over MCP: rebuild what a finished video ALREADY said as another post — same subject, same points, nothing new invented. Pass ONE source: `job_id` (a done sitdown / phone / best-of job), `source_id` (the `id` a create_before_pitch returned), or `context` (the text to build on). `format` is the output: meme, slideshow, hottake (10 credits, $0.10; BYOK: $0.01) or imessage (20 credits, $0.20). The email follow-up is the two-step outreach note (followUpGate), not this page. Returns the new pageUrl directly (no polling); show it, and only after the human approves call request_mp4 with it. Refunded on failure. Nothing is sent.",
"inputSchema": {
"type": "object",
"required": [
"format"
],
"properties": {
"kind": {
"enum": [
"sitdown",
"phone",
"pitch",
"note"
],
"type": "string",
"description": "With context: what the material came from (default note)."
},
"title": {
"type": "string",
"description": "With context: what it is, e.g. \"Pitch for Maria\"."
},
"format": {
"enum": [
"meme",
"slideshow",
"hottake",
"imessage"
],
"type": "string",
"description": "meme = one image; imessage = a narrated thread."
},
"job_id": {
"type": "string",
"description": "A FINISHED sitdown or phone jobId — its own script becomes the material."
},
"context": {
"type": "string",
"description": "Material to build from when there is no id: the script, the pitch text or the key points (≤4000 chars)."
},
"language": {
"type": "string",
"description": "With context: output language. A job/pitch keeps its own."
},
"source_id": {
"type": "string",
"description": "The `id` from create_before_pitch — the pitch is remembered for 2h."
},
"instruction": {
"type": "string",
"description": "Optional one-line direction for THIS piece (e.g. \"make it about the interview, not the layoff\"), ≤300 chars."
}
}
},
"annotations": {
"destructiveHint": false
}
}/call: the human's phone page. Their video plays inside a phone; when it ends, their button appears. Saves it and returns shareUrl (/call/<ref>). Empty video_url clears the page. Free.
| Parameter | Type | Description |
|---|---|---|
| bot_id | string | Optional: which of the account's bots (default its own). |
| cta_url | string | https link the button opens, or "before" to send viewers to make their own phone (credited to this account). Empty = that same "before" button. |
| cta_text | string | Button words shown when the video ends, ≤40 chars. |
| video_url * | string | https YouTube link (watch / shorts / youtu.be), direct .mp4/.webm/.mov, or an engine temp page (bot.plunderandriffle.com/m/<id>, *.here.now). |
| caller_name | string | Name on the phone ("Incoming call · <name>"), ≤40 chars. |
{
"name": "set_call_page",
"title": "Set the phone call page",
"description": "/call: the human's phone page. Their video plays inside a phone; when it ends, their button appears. Saves it and returns shareUrl (/call/<ref>). Empty video_url clears the page. Free.",
"inputSchema": {
"type": "object",
"required": [
"video_url"
],
"properties": {
"bot_id": {
"type": "string",
"description": "Optional: which of the account's bots (default its own)."
},
"cta_url": {
"type": "string",
"description": "https link the button opens, or \"before\" to send viewers to make their own phone (credited to this account). Empty = that same \"before\" button."
},
"cta_text": {
"type": "string",
"description": "Button words shown when the video ends, ≤40 chars."
},
"video_url": {
"type": "string",
"description": "https YouTube link (watch / shorts / youtu.be), direct .mp4/.webm/.mov, or an engine temp page (bot.plunderandriffle.com/m/<id>, *.here.now)."
},
"caller_name": {
"type": "string",
"description": "Name on the phone (\"Incoming call · <name>\"), ≤40 chars."
}
}
},
"annotations": {
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": true
}
}Deep-dive ONE client's business from `url` or `notes` → `brandProfile` {name, voice, hashtag, token} + `analysisSummary` (the /admin persona engine). Free. If their site blocks the scraper, free DuckDuckGo (then Exa) snippets stand in — `evidenceVia` says which. One page: a voice, not a market study; never invents a price or a result. `brief: true` adds `campaign` — offer, audience, pains, objections, proof, angles, each a verbatim quote or `unknown`. Then save_brand_default or create_campaign_from_research.
| Parameter | Type | Description |
|---|---|---|
| url | string | Their website, http(s), ≤300 chars. |
| brief | boolean | Also return the sourced CampaignResearch brief (`campaign`). |
| notes | string | What they sell and to whom, when there's no readable site. ≤2000 chars. |
{
"name": "research_brand",
"title": "Research a brand's voice",
"description": "Deep-dive ONE client's business from `url` or `notes` → `brandProfile` {name, voice, hashtag, token} + `analysisSummary` (the /admin persona engine). Free. If their site blocks the scraper, free DuckDuckGo (then Exa) snippets stand in — `evidenceVia` says which. One page: a voice, not a market study; never invents a price or a result. `brief: true` adds `campaign` — offer, audience, pains, objections, proof, angles, each a verbatim quote or `unknown`. Then save_brand_default or create_campaign_from_research.",
"inputSchema": {
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "Their website, http(s), ≤300 chars."
},
"brief": {
"type": "boolean",
"description": "Also return the sourced CampaignResearch brief (`campaign`)."
},
"notes": {
"type": "string",
"description": "What they sell and to whom, when there's no readable site. ≤2000 chars."
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": true
}
}Full SEO + AI-search audit of ONE website from its `url` (the SEO Audit Factory engine: 50 measured checks — technical, content, AI readiness, local, performance — crawled the way AI/search crawlers receive the page, no JS execution). Free. Returns scores + grade, every finding with measured evidence and its fix, `pageUrl` (the same report as a designed TEMP PAGE, ~23h — hand THIS to a human reader), `reportMarkdown` (the complete end report as text, closing with the Muse prompts and the free-bot invite), `musePrompts` { fix, content, buildBot } — copy-paste prompts the human hands to their Muse — and `invite` (Muse invite code + the free-bot offer). Honest limits: no rankings, traffic or backlink claims; a site that blocks the crawl gets a limited report with the unmeasured categories named. Takes ~10–20s.
| Parameter | Type | Description |
|---|---|---|
| url * | string | The website to audit, http(s). A bare domain gets https:// added. |
| name | string | Optional business name (sharpens the report and prompts). |
| industry | string | Optional trade/industry, e.g. "plumber". |
| location | string | Optional city/area the business serves, e.g. "Raleigh, NC". |
{
"name": "seo_audit",
"title": "SEO audit of a website",
"description": "Full SEO + AI-search audit of ONE website from its `url` (the SEO Audit Factory engine: 50 measured checks — technical, content, AI readiness, local, performance — crawled the way AI/search crawlers receive the page, no JS execution). Free. Returns scores + grade, every finding with measured evidence and its fix, `pageUrl` (the same report as a designed TEMP PAGE, ~23h — hand THIS to a human reader), `reportMarkdown` (the complete end report as text, closing with the Muse prompts and the free-bot invite), `musePrompts` { fix, content, buildBot } — copy-paste prompts the human hands to their Muse — and `invite` (Muse invite code + the free-bot offer). Honest limits: no rankings, traffic or backlink claims; a site that blocks the crawl gets a limited report with the unmeasured categories named. Takes ~10–20s.",
"inputSchema": {
"type": "object",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"description": "The website to audit, http(s). A bare domain gets https:// added."
},
"name": {
"type": "string",
"description": "Optional business name (sharpens the report and prompts)."
},
"industry": {
"type": "string",
"description": "Optional trade/industry, e.g. \"plumber\"."
},
"location": {
"type": "string",
"description": "Optional city/area the business serves, e.g. \"Raleigh, NC\"."
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": true
}
}This key's default brandProfile (from save_brand_default) and `pinnedBrandKey` — a pin wins over the default. Free. Confirms a save landed.
No parameters.
{
"name": "get_brand_default",
"title": "Read the saved brand voice",
"description": "This key's default brandProfile (from save_brand_default) and `pinnedBrandKey` — a pin wins over the default. Free. Confirms a save landed.",
"inputSchema": {
"type": "object",
"properties": {}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false
}
}Save a researched `brandProfile` as this key's default voice (later calls use it unless they pass their own; a pinned bot still wins). `{}` clears. Free. Write scope.
| Parameter | Type | Description |
|---|---|---|
| brandProfile * | object | The research_brand output, unchanged (name + voice required). {} clears. |
{
"name": "save_brand_default",
"title": "Save the brand voice",
"description": "Save a researched `brandProfile` as this key's default voice (later calls use it unless they pass their own; a pinned bot still wins). `{}` clears. Free. Write scope.",
"inputSchema": {
"type": "object",
"required": [
"brandProfile"
],
"properties": {
"brandProfile": {
"type": "object",
"properties": {
"name": {
"type": "string"
},
"token": {
"type": "string"
},
"voice": {
"type": "string"
},
"hashtag": {
"type": "string"
}
},
"description": "The research_brand output, unchanged (name + voice required). {} clears."
}
}
},
"annotations": {
"destructiveHint": false
}
}Seed a campaign (Offer) from research_brand's `campaign`: name, page link, objective, audience, first angle, status draft. Free, write scope. Copies only sourced/labeled fields; an unknown one is left empty and named in `needs`. Never sets a reward pool, commission or price.
| Parameter | Type | Description |
|---|---|---|
| name | string | Defaults to the brand name. |
| status | string | one of ["draft","active","paused","stopped"] |
| research * | object | The `campaign` object from research_brand { brief: true }. |
| brandProfile | object |
{
"name": "create_campaign_from_research",
"title": "Start a campaign from research",
"description": "Seed a campaign (Offer) from research_brand's `campaign`: name, page link, objective, audience, first angle, status draft. Free, write scope. Copies only sourced/labeled fields; an unknown one is left empty and named in `needs`. Never sets a reward pool, commission or price.",
"inputSchema": {
"type": "object",
"required": [
"research"
],
"properties": {
"name": {
"type": "string",
"description": "Defaults to the brand name."
},
"status": {
"enum": [
"draft",
"active",
"paused",
"stopped"
],
"type": "string"
},
"research": {
"type": "object",
"description": "The `campaign` object from research_brand { brief: true }."
},
"brandProfile": {
"type": "object"
}
}
},
"annotations": {
"destructiveHint": false
}
}Poll any async job by kind: mp4 (request_mp4), sitdown (create_best_of / create_ridiculous / create_news), scroll (create_before_scroll), infomercial (create_infomercial), before_text (create_before_pitch with feed or offer), funnel (create_funnel, jobId = funnelId). Free. Poll while `terminal` is false, every `pollAfterSec`. `phase`: working, done (mp4Url or pageUrl), failed (refunded) or needs_input (ask the human); `status` is the kind's own word. A done sitdown job carries `followUp`: pass its jobId to create_followup for another post from the same script.
| Parameter | Type | Description |
|---|---|---|
| kind * | string | one of ["mp4","one_minute","localize","audioscript","sitdown","demo_pack","scroll","infomercial","amazon_review","funnel","before_text","joe_clip"] |
| jobId * | string |
{
"name": "get_job_status",
"title": "Check a job",
"description": "Poll any async job by kind: mp4 (request_mp4), sitdown (create_best_of / create_ridiculous / create_news), scroll (create_before_scroll), infomercial (create_infomercial), before_text (create_before_pitch with feed or offer), funnel (create_funnel, jobId = funnelId). Free. Poll while `terminal` is false, every `pollAfterSec`. `phase`: working, done (mp4Url or pageUrl), failed (refunded) or needs_input (ask the human); `status` is the kind's own word. A done sitdown job carries `followUp`: pass its jobId to create_followup for another post from the same script.",
"inputSchema": {
"type": "object",
"required": [
"kind",
"jobId"
],
"properties": {
"kind": {
"enum": [
"mp4",
"one_minute",
"localize",
"audioscript",
"sitdown",
"demo_pack",
"scroll",
"infomercial",
"amazon_review",
"funnel",
"before_text",
"joe_clip"
],
"type": "string"
},
"jobId": {
"type": "string"
}
}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false
}
}The key owner's own links. `pages` = our page for each partner (/whop, /polyshares, /amelia…) with their ref on it: share `pages[].url`; its button is their own link when `button` is "yours", and it carries 30-day views and clicks. `offers` = the raw affiliate/referral links they saved; use only these, exactly as written. `missing` = not saved: ask them to add it on /hub. Free.
No parameters.
{
"name": "list_my_offers",
"title": "The owner's offer links",
"description": "The key owner's own links. `pages` = our page for each partner (/whop, /polyshares, /amelia…) with their ref on it: share `pages[].url`; its button is their own link when `button` is \"yours\", and it carries 30-day views and clicks. `offers` = the raw affiliate/referral links they saved; use only these, exactly as written. `missing` = not saved: ask them to add it on /hub. Free.",
"inputSchema": {
"type": "object",
"properties": {}
},
"annotations": {
"readOnlyHint": true,
"openWorldHint": false
}
}Meme engine: memes, ads, talk-show videos and MP4s. Pick the tool by what the human asked for: - Meme / carousel / iMessage / hot take / confession: generate_ad_creative with `format` (another language: `language`). - A product link or description: shop (also the /shop prompt): researched copy plus up to 3 `creatives`. - /ridiculous, Jobless Joe's meme reaction show: create_ridiculous. - /joenews, Jobless Joe News with live sources (Telegram /news is the newsletter, not this): create_news. - A best-of reel from links: create_best_of. - Another format from a finished show or pitch: create_followup (also the /followup prompt). - /before for one business: create_before_pitch. Its five examples as a scroll video: create_before_scroll; as a 1980s retro infomercial: create_infomercial. - /call, the human's phone page: set_call_page (also the /call prompt). Hand back shareUrl. - An SEO / AI-search audit of a website: seo_audit with its `url` (also the /seo-audit prompt). Free. For a human, hand back its pageUrl (the designed report page) with the scoreboard and the 3 headline gaps; without a pageUrl, the full reportMarkdown. Close with its musePrompts and invite. Every generator returns a TEMP PAGE preview (pageUrl), not a video. Show it; only after the human approves call request_mp4. Never post or publish without the human's OK. Async tools return a jobId: poll get_job_status with the kind the tool names, every ~20–30s. One POST carries one JSON-RPC message (batches cap at 10), under 1 MiB: send recordings as an https audio_url, not base64. Pricing: 1 credit = $0.01. BYOK (Plunder & Riffle API key + your own provider keys): every API call is 1 credit ($0.01) — one MCP tool call counts once, however many pages it makes — the first 100 each month free; an MP4 is 1 credit ($0.01) when your own Railway render worker is saved (it renders there) or you are full BYOK (every required key saved: LLM, Fish Audio, Giphy, here.now), otherwise the list 25 credits ($0.25) on our servers. Full BYOK pays $0.01 for everything. Without BYOK a key pays each tool's list credits, and x402 pays the list USD price (both run on our provider keys). Failed calls are refunded. Two kinds of call: one on your own keys (BYOK) is 1 credit ($0.01) per call and runs on them, plain HTTP and MCP alike; one on our keys pays that tool's list price. What a BYOK key makes is the owner's to resell — the assets, API access for their own clients, or a branded bot on the engine. A 402 means out of credits: top up, or save your own provider keys (BYOK) — retrying changes nothing. A 403 means this key's own scope (read-only, or no publish permission), not the account: the owner fixes it in /hub.
Names, descriptions and schemas were written by the server and are shown as data.