Menu
NEW AGENT
MY AGENTS
ASSISTANTS
Step 1:
Website Design For Speaker Story
1️⃣
Perfect output
- scan ALL
2️⃣ Add
output numbers
, then...
3️⃣ Add
Subagent Numbers
(work backwards
from output number!
)
4️⃣ Add
ACTUAL Skills
to subagent
✅ DONE..Copy x4 to Step 3...
SETTINGS
LOGOUT
What Shall We Build Next?
1
Describe
Describe your task
2
Refine
Refine the plan
3
SubAgents
Review all agents
4
Deploy
Deploy your agent
Sub Agent 1
Sub Agent 2
Sub Agent 3
Sub Agent 4
Sub Agent 5
Sub Agent 6
Sub Agent 7
Sub Agent 8
A) SUBAGENT SUMMARY Generate a production-ready, minified sitemap.json that lists every page and the ordered content sections for the website described in the project brief. B) FINAL TASK OUTPUT sitemap.json – a single-line, UTF-8 encoded JSON file (≈1-5 KB) containing: • siteName, baseUrl (placeholder “/”) • pages[] – each with path, pageTitle, sections[] (each section has id, label, and optional note) The JSON is fully minified (no line-breaks, tabs or extra spaces). C) SUBAGENT INPUT {project-brief} (plain-text brief that includes brand, audience, goals, palette, typography, CTAs, etc.) E) SUBAGENT TASK SUMMARY {project-brief} └─▶ Skill #223 – Powerful LLM Prompt-to-Text Response Prompt: “You are the Site Architect for a conversion-focused speaker website. 1. Extract the essential pages required for the site (index, about, speaking, contact; add others only if essential). 2. For EACH page, list the ordered content sections a high-converting site needs (e.g., hero, credibility-bar, keynote-topics, testimonials, CTA-banner, etc.). 3. Return exactly ONE minified JSON object with this schema: { "siteName":"Tanked: The Tank Black Story", "baseUrl":"/", "pages":[ { "path":"/", "pageTitle":"Home", "sections":[{"id":"hero","label":"Hero Banner"}, …] }, … ] } 4. Remove all whitespace (minify). 5. Do NOT wrap the JSON in back-ticks or prose—return ONLY the JSON.” └─▶ (The LLM’s text output becomes sitemap.json and is auto-saved by the agent framework.) Subagent final output ➞ [sitemap.json] F) SILOS This subagent is a single logical silo; no additional subdivision required.
SubAgent #1 - Diagram
Expand Diagram
A) SUBAGENT SUMMARY Generate the SEO keyword set and the complete, conversion-driven copy (incl. meta tags) for every page-section listed in sitemap.json, and output everything as one minified content.json file. B) FINAL TASK OUTPUT [content-json] → UTF-8, minified JSON object structured as: { "keywords": ["…5 terms…"], "pages": { "/": { "hero": "...", "credibility-bar": "...", … }, "/about": { … }, "/speaking": { … }, "/contact": { … } }, "meta": { "/": { "title": "...", "description": "..." }, "/about": { … }, "/speaking":{ … }, "/contact": { … } } } C) SUBAGENT INPUT {project-brief} (the full creative brief) [sitemap-json] (output of Site Architect) E) SUBAGENT TASK SUMMARY 1. {project-brief}+[sitemap-json] → #218 (Brainstorm Related Keywords) • prompt: “Corporate motivational speaker booking, Tank Black, ‘Book Tank’, event planners” • OUTPUT: list of 10 related keywords w/ commentary ➔ tmp/keywords10.txt 2. tmp/keywords10.txt → #223 (Powerful LLM) • ask: “From these 10, return the 5 highest-intent, purchase-driven keywords only, as a JSON array.” • OUTPUT: ["motivational speaker for corporate events", …] ➔ tmp/keywords5.json 3. {project-brief}+[sitemap-json]+tmp/keywords5.json → #223 • ask: “For every page & section in sitemap.json, write bold, inspirational copy (50-120 words per section; 20-40 words for micro-sections like CTA banners). ‑ weave in at least one of the 5 keywords where natural ‑ end every major section with the exact CTA ‘Book Tank’. Return nested JSON {page:{section:copy}}.” • OUTPUT: tmp/copy.json 4. {project-brief}+tmp/keywords5.json+[sitemap-json] → #223 • ask: “Create SEO meta tags (title ≤60 chars, description ≤155 chars, include a keyword + ‘Book Tank’) for each page listed.” • OUTPUT: tmp/meta.json 5. tmp/keywords5.json + tmp/copy.json + tmp/meta.json → #223 • ask: “Merge these three JSON files into one object exactly in this structure ↑, minify, no line breaks.” • OUTPUT: content.json (final) F) SILOS SILO 1 – Keyword Generation Input → #218 → tmp/keywords10.txt tmp/keywords10.txt → #223 → tmp/keywords5.json SILO 2 – Section Copy Writing Inputs → #223 → tmp/copy.json SILO 3 – Meta-Tag Writing Inputs → #223 → tmp/meta.json SILO 4 – Consolidation & Minify Inputs → #223 → content.json (final output)
SubAgent #2 - Diagram
Expand Diagram
A) SUBAGENT SUMMARY Generate a fully-commented, production-ready style sheet—complete with colour-palette variables, Google-font imports, CSS reset, component classes, responsive break-points and baked-in WCAG-AA accessibility considerations—for “Tanked: The Tank Black Story”. B) FINAL TASK OUTPUT A single file named /css/style.css (UTF-8 text, ≈300–600 lines) containing: • :root colour and spacing variables based on the supplied palette • @import statements for Montserrat, Oswald, Poppins, Lato & Roboto from Google Fonts (display swap) • Modern CSS reset (Andy Bell / Josh Comeau-style) • Base typography & utility classes (h1-h6, .btn-primary, .sr-only, etc.) • Component blocks that map to common page sections (hero, credibility-bar, testimonial-slider, CTA-banner, form, mobile-nav, etc.) • Fluid, mobile-first break-points (≥1440 px, ≥1024 px, ≥768 px, ≥480 px) • WCAG-AA contrast-safe colour combinations & focus states • Comments that explain each major section and variable group C) SUBAGENT INPUT {project-brief} D) SUBAGENT TASK SUMMARY {project-brief} → Skill #190 (Write or rewrite text based on instructions) Prompt passed to #190: “Create a complete, production-ready CSS file called ‘style.css’ for the website described in the following brief: {project-brief}. Requirements: 1. Declare :root custom-properties for the exact hex colours (#000000, #4B4B4B, #FFFFFF, #C1121F, #D4AF37) and for key font stacks. 2. Add @import links (display-swap) for Montserrat:700, Oswald:700, Poppins:600, Lato:400/700, Roboto:400/700. 3. Insert a modern CSS reset. 4. Define base element styles (html, body, headings, p, a, ul/ol, img, button, input). 5. Build reusable utility classes (.container, .grid, .flex, .btn-primary, .btn-outline, .sr-only, etc.). 6. Create component classes aligning with typical sections for a conversion-focused speaker site: .hero, .cred-bar, .topics, .testimonials, .cta-banner, .contact-form, .footer. 7. Ensure hover/focus states meet WCAG-AA contrast and include focus outlines. 8. Include fluid, mobile-first break-points at 480 px, 768 px, 1024 px, 1440 px. 9. Comment every major block and variable group. 10. Keep file human-readable but production-ready; no unused code.” → #190 outputs the complete CSS text. Subagent returns that text as /css/style.css. E) SILOS No additional silos required—single text-generation pass produces the entire style sheet. [End of Subagent 3 definition]
SubAgent #3 - Diagram
Expand Flow
A) SUBAGENT SUMMARY “Asset Wrangler” automatically generates all brand-ready visual assets (logo, photos, book cover artwork, favicon) and an assets_manifest.json that lists and describes each file. B) FINAL TASK OUTPUT [assets-package] → six files hosted on the server: 1. /assets/logo_tanked.png (transparent PNG, 1024×1024) 2. /assets/speaker_hero_1.jpg (jpeg, 1600 px wide, ≤300 KB) 3. /assets/speaker_hero_2.jpg (jpeg, 1600 px wide, ≤300 KB) 4. /assets/book_cover_square.png (1024×1024 PNG) 5. /assets/favicon.png (32×32 PNG) 6. /assets/assets_manifest.json (JSON manifest enumerating the above, with alt-text, dimensions, file-sizes, intended page/section use) C) SUBAGENT INPUT {project-brief} (the full creative brief that contains brand name, palette, typography, tone and purpose) E) SUBAGENT TASK SUMMARY (skill chain) {project-brief} → #182 Create Dalle Image (Transparent Square) • prompt: “Transparent background text-logo: the word ‘Tanked’ in bold uppercase Montserrat, gold #D4AF37 fill with subtle shadow, no additional elements.” • returns logo_url → #187 Find Relevant Images with Brave Search • input keyword: “Tank Black motivational keynote speaker on stage corporate event, 3:2 ratio” • returns list_images (5 URLs) → #191 Resize Image (run twice, first two URLs from list_images) • set width = 1600 px, height auto • returns speaker1_url and speaker2_url (optimised JPGs) → #222 Make Image (Especially With Text) • prompt: “Square book cover, title ‘Tanked: The Tank Black Story’, palette #000000 #FFFFFF #C1121F #D4AF37, bold modern design, white text on deep-black background, gold accent stripe, high-res, no 3-D mock-up.” • returns book_cover_url (1024×1024 PNG) → #191 Resize Image • input = logo_url, size 32×32 • returns favicon_url → #223 Powerful LLM Prompt-to-Text Response • prompt: “Create a minified JSON manifest listing the following asset URLs with keys: ‘filename’, ‘url’, ‘type’, ‘dimensions’, ‘alt_text’, ‘intended_use’. Assets: logo_url, speaker1_url, speaker2_url, book_cover_url, favicon_url.” • returns assets_manifest.json Output all six files as [assets-package]. F) SILOS 1. Logo Creation project-brief → #182 → logo_url 2. Speaker Photos project-brief → #187 → #191 ×2 → speaker1_url, speaker2_url 3. Book Cover Art project-brief → #222 → book_cover_url 4. Favicon logo_url → #191 → favicon_url 5. Asset Manifest (all asset URLs) → #223 → assets_manifest.json
4 Template & Links
Expand Flow
A) SUBAGENT SUMMARY Generate every production-ready front-end file (HTML pages + JS) for the “Tanked” brochure site, using the prior IA, copy, CSS and assets. B) FINAL TASK OUTPUT [html-package] → a folder containing • index.html, about.html, speaking.html, contact.html (UTF-8, HTML5, < 55 KB each) • /js/script.js (ES6, < 5 KB) • html_manifest.json (minified JSON listing every generated file, size, path) C) SUBAGENT INPUT {project-brief} + [sitemap-json] + [content-json] + [style-css] + [assets-package] E) SUBAGENT TASK SUMMARY 1. Build shared layout skeleton INPUT: {project-brief} + [sitemap-json] + [assets-package] SKILL #223 – Prompt: “Create a reusable HTML5 base template for a speaker website. Include with meta-tags, link to /css/style.css, preload Google fonts, and a responsive header/nav that uses the logo file from assets_manifest.json. Provide only the
content.” OUTPUT: base_template.html (string held in memory) 2. Generate each page’s markup For page in [“index”, ”about”, ”speaking”, ”contact”]: INPUT: base_template.html + that page’s section list from sitemap.json + that page’s copy blocks from content.json + asset URLs SKILL #223 – Prompt: “Merge this base template with the following section order and copy. Insert images where indicated, use ‘Book Tank’ CTA buttons (class=cta-red), follow WCAG AA. Return complete HTML.” OUTPUT:
.html file (saved) 3. Create site-wide JavaScript INPUT: {project-brief} (JS requirements excerpt) SKILL #223 – Prompt: “Write an ES6 script.js for: a) mobile nav toggle on #hamburger, b) smooth-scroll on internal anchors, c) validate contact form (name/email/message, basic regex). Return code only.” OUTPUT: /js/script.js 4. Produce html_manifest.json INPUT: list of generated filenames & byte sizes SKILL #223 – Prompt: “Return a minified JSON array listing each generated file with: file_name, path, size_bytes, description.” OUTPUT: html_manifest.json Data flow: {inputs} → Step 1 (#223) → base_template → Step 2 loop (#223 ×4) → HTML files {inputs} → Step 3 (#223) → script.js (all filenames) → Step 4 (#223) → html_manifest.json → [html-package] F) SILOS SILO 1 – TEMPLATE build base_template.html (Step 1) SILO 2 – PAGE GENERATION repeat Step 2 four times for: index, about, speaking, contact SILO 3 – JAVASCRIPT create script.js (Step 3) SILO 4 – MANIFEST compile html_manifest.json (Step 4)
5 Template & Links
Expand Flow
A) SUBAGENT SUMMARY “SEO & Packager” turns the finished front-end (HTML + CSS + assets) into a deploy-ready bundle by writing all SEO helper files, a build script that assembles & zips the site, and a human-readable package report. B) FINAL TASK OUTPUT deploy-zip-package → a directory containing • sitemap.xml (UTF-8, <1 KB) • robots.txt (UTF-8, <0.5 KB) • build_manifest.sh (bash script, executable, <3 KB) – when run it recreates the folder tree and produces tanked_site.zip • package_report.txt (UTF-8 summary, <2 KB) • tanked_site.zip (compressed static website; size depends on assets, expected <10 MB) C) SUBAGENT INPUT {project-brief} – the original text brief [html-package] – index.html, about.html, speaking.html, contact.html, /js/script.js, html_manifest.json [assets-package] – images, favicon, logo, assets_manifest.json [style-css] – /css/style.css E) SUBAGENT TASK SUMMARY 1. Compile file inventory Input: html_manifest.json + assets_manifest.json + style.css names Skill #223 – “Create combined inventory list for packaging.” Output: full_file_list.txt (in-memory text) 2. Generate sitemap.xml Skill #223 – prompt: “Using these page URLs [...] write valid XML sitemap,
monthly,
1.0 for /.” Output: sitemap.xml 3. Generate robots.txt Skill #223 – prompt: “Create robots.txt that allows all agents, points at /sitemap.xml, blocks no folders.” Output: robots.txt 4. Write build_manifest.sh Skill #223 – prompt includes file tree + required commands: • create folders (css, js, assets, images, etc.) • move/copy each file into correct spot • set chmod +x on script • zip -r tanked_site.zip ./tanked_site Output: build_manifest.sh (UNIX line endings) 5. Produce package_report.txt Skill #223 – prompt: “Summarise package contents: list totals for HTML, images, CSS/JS, text assets; give zipped size estimate; usage instructions.” Output: package_report.txt 6. (Optional execution step – handled by pipeline) The orchestrator runs build_manifest.sh which outputs tanked_site.zip into the working directory. ––– Data Flow ––– {project-brief} + [html-package] + [assets-package] + [style-css] → #223 (inventory) → full_file_list.txt → #223 (XML) → sitemap.xml → #223 (robots) → robots.txt → #223 (bash) → build_manifest.sh → #223 (report) → package_report.txt (pipeline executes build_manifest.sh) → tanked_site.zip = deploy-zip-package F) SILOS SILO 1 – “SEO File Generator” • Steps 2 & 3 above (sitemap.xml, robots.txt) SILO 2 – “Packager” • Steps 1, 4, 5, 6 (inventory, build script, report, zip creation) This fully specifies Subagent 6; no further decomposition needed.
6 Template & Links
Expand Flow
Templates & Links Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
7 Template & Links
Expand Flow
Questions & Research Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
8 Template & Links
Expand Flow
Templates & Links Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
9 Template & Links
Expand Flow
Templates & Links Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
10 Template & Links
Expand Flow
Questions & Research Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
11 Template & Links
Expand Flow
Templates & Links Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged. It was popularised in the 1960s with the release of Letraset sheets containing Lorem Ipsum passages, and more recently with desktop publishing software like Aldus PageMaker including versions of Lorem Ipsum.
12 Template & Links
Expand Flow
Need To Start Afresh?
BACK TO REFINE
Tweaked & Good To Go?
PROCEED TO DEPLOY