Search the docs, videos, blog, apps, teams, skills and release notes.
At a glance
What it is
Strip modules that do not earn a measurable lift. Every block must answer a buyer question.
A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.
Each of these library teams gives its bots this skill. Import one and look inside first.
The whole file, exactly as it downloads. Read it before you rely on it.
SKILL.md
89 lines · 801 words
As of: 2026-05-16
Mode skill. Default-enabled on the Storefront specialist.
Use when the user is building, auditing, or rebuilding the brand-owned storefront. Use when the brief involves Shopify (or comparable) setup, PDP architecture, collection structure, or conversion-rate fundamentals. Use before discussing marketplaces or agentic surfaces — those skills assume a healthy home base.
Trigger phrases:
Strip modules that do not earn a measurable lift. Every block must answer a buyer question.
Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.
Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.
Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.
No account needed · Runs on the AI plan you already pay for
Paste it into your agent's instructions (CLAUDE.md, AGENTS.md or a custom GPT), or add it to a bot in Murage.
Ask Ember for the bot, then switch the skill on in its settings.
Step 1
Send your Chief of Staff this message. It proposes the bot and its role first, and creates it when you say go.
Send this to Ember
“Create a bot that uses the Storefront Foundation skill, and show me the plan first.”
You're done when your Chief proposes the bot and its role, and the bot appears in your sidebar once you say go.
Step 2
You're done when you see the bot’s skills and the library.
Step 3
You're done when the skill shows as on for that bot.
Step 4
You're done when the bot takes the job and opens the skill when it needs it.
Prefer to do it yourself? Choose + in the sidebar, then New Bot.Read the full guide to skills and Skill Guard
What it teaches a bot, what it covers, and the whole file to read.
The principle
**Mode skill.** Default-enabled on the Storefront specialist.
What it covers
When to use
Procedure
Decision rules
Anti-patterns
Before / after
"Audit our Shopify store."
"Help me design a product detail page."
"Our conversion rate is X% — what's wrong?"
"We're launching a new product line."
If Beacon has posted traffic-source breakdowns to TEAM_MEMORY.md, start at step 2.
1. Pull the three numbers. Before any design move: site-wide conversion rate, average order value, primary traffic source by share. If the user does not have them, the deliverable is "go get these three numbers" — refuse to design blind.
2. Diagnose the conversion fundamentals. Independent of traffic source, walk the page on mobile in airplane mode and grade six fundamentals:
Speed: LCP under 2.5s, INP under 200ms. Anything slower is a conversion tax.
Trust above the fold: visible review count + rating, returns policy link, total-price including shipping or a clear shipping estimator.
Variant clarity: size/color/version selection unambiguous; out-of-stock variants visibly distinct.
Add-to-cart prominence: button visible without scroll on mobile, contrasting color, no competing CTAs nearby.
Cart and checkout: one-page or short-step checkout, guest checkout offered, payment methods matched to market (Apple Pay, Shop Pay, local wallets), cart abandonment recovery wired up.
Imagery: hero shot in-context, alternate angles, scale reference, video for high-consideration SKUs.
Score each red/yellow/green. Fix all reds before any other work.
3. Architect the PDP module stack. Decide the order and presence of modules:
Hero (image + title + price + ATC + trust strip)
Variant selector
Short benefit bullets (3-5)
Lifestyle / problem-solution block
Spec / fit / sizing block
Reviews (with photos, sortable)
FAQ (questions tied to real CS tickets)
Comparison table (if category is comparison-driven)
Cross-sell / bundle block
Strip modules that do not earn a measurable lift. Every block must answer a buyer question.
4. Set the collection and navigation structure. Collections map to how the buyer shops (by use-case, by problem, by price tier), not how you organize inventory internally. Navigation depth: two clicks to any product. Search bar visible on mobile; site search results must be tunable.
5. Define the measurement loop. Per-SKU: PDP conversion rate, add-to-cart rate, cart-to-checkout rate, return rate, review velocity. Site-wide: revenue per session, AOV, repeat-purchase rate. Hand the tracking spec to whoever owns analytics infrastructure.
Three numbers or no design. Conversion rate, AOV, traffic source. Never start without them.
Mobile-first, no exceptions. 70%+ of DTC traffic is mobile. Design mobile, sanity-check desktop.
One promise per page. A PDP that tries to convert three different buyer personas converts none. Pick the dominant buyer; build for them.
Reviews are infrastructure, not garnish. Below 20 reviews, social proof is fragile. Above 100, it is a moat. Wire review-request emails before launch day, not after.
Speed is conversion. A 1-second mobile LCP improvement typically lifts conversion 5-15%. Spend the engineering budget here before redesigning anything visual.
Redesigning the PDP because the team is bored of looking at it. The buyer has never seen it.
Hero video that auto-plays with sound — bounces buyers on mobile.
"Hero carousels" that nobody clicks past slide one. Pick one promise.
Stacking apps that each add 200ms of load time until the page is a slideshow.
Building collections that mirror your warehouse instead of the buyer's mental model.
Hiding shipping cost until step three of checkout. The buyer will leave when they find it.
Treating PDP copy as "set and forget." Real CS tickets and search queries change quarterly — refresh the FAQ and bullets against them.
Brief: "Our Shopify store is converting badly — redesign it."
Before (blind redesign):
Mock up a new homepage with a video hero and a cleaner product grid.
After (numbers-first, traffic-source-aware, 2026-05-16):
Step one: pull conversion (2.1%), AOV ($68), dominant source (Meta paid = 64%). Step two: audit fundamentals — mobile LCP 4.1s (red), no shipping estimator above fold (red), reviews below fold on mobile (yellow). Step three: with Meta-paid dominant, the PDP is the landing page — rebuild hero to continue ad creative within first scroll, move reviews above fold, fix LCP via image compression and app audit. Measure: PDP conversion at 7 / 14 / 30 days; target +25% within 30 days from speed and trust placement.
SKILL.md
89 lines · 801 words
The text of SKILL.md is shown as it downloads, under the Apache-2.0 license, by Ferrox Labs. The license is the one in the file's own frontmatter.
Your new bot
Storefront Foundation bot
Waiting in your sidebar once you say go.
No account needed · Runs on the AI plan you already pay for