# Scrappy Start, full content > The full text of every public page and skill on scrappystart.ai, concatenated for agents. > Generated dynamically from the same source files that build the human-facing site. > Last fetched: 2026-05-19T00:23:56.190Z Scrappy Start teaches small business owners how to use Claude (Anthropic's AI) on the real problems in their business. Live 1:1 sessions on Google Meet, $150 a session, first one free, no subscription. Run by Jamie and Erica Sowder. Site: https://scrappystart.ai Email: jamie@scrappystart.ai License: skills below are CC0. Site copy is © Scrappy Start; please cite when summarizing. --- ## Page: home (https://scrappystart.ai/) Headline: Cancel the SaaS. Claude already does it. You just need the skill. I sit with you for an hour and we use Claude (the AI from Anthropic) to solve one real thing in your business. The inbox at 9pm. The quote you keep putting off. The Etsy order you've answered forty times. Whatever ate your week. I start tackling it while you watch how I think. Mid-way we trade off, hands on your keyboard. You finish it before next session. Next time we review, sharpen what worked, and pick the next thing. After a few months you don't need me, your skill keeps compounding, and your evenings come back. The shape of an hour ("I do, we do, you do"): 1. Bring one real thing. A bid you're tired of writing. An email you keep putting off. A spreadsheet that takes you forty minutes. One thing. 2. I do. I share my screen and solve it in front of you in Claude. You see the prompts. You see when I'm wrong and how I fix it. 3. We do. Your turn at the keyboard. Your screen, your data. I'm in your ear. You drive, I coach. 4. You do. You keep at it on your own real work. Reps. That's where the muscle gets built. 5. Next session, we review what you tried, what worked, what got weird. Then we go to the next thing. Categories of work: customer email triage, the questions you've answered a hundred times, orders and quotes and back-and-forth, captions and posts, booking and follow-up, the one weird process only your business has. ## Page: the arrangement (https://scrappystart.ai/the-arrangement) What an hour looks like: 60-90 minutes, live, on Google Meet, both screens shared. You bring one real piece of work. I show you how I'd handle it in Claude. Then you do the next one with my voice in your ear. You leave the call with the thing actually done and a move you can repeat. What it costs: $150 a session. First session is free. Book as many as you want, when you want. Most people come weekly while building the muscle. No subscription, no contract, no setup fee. If a session doesn't deliver value, refund. What you bring: A Claude subscription paid to Anthropic. Pro ($20/mo) is the prereq, you'll likely want Max ($100 or $200/mo) within a few months. Plus one real piece of work each session. What you own: Everything we build runs on your account. You own the prompts, the setups, the moves. The skill stays with you. What we won't do: Run your business for you. Ship work under your name without you reading it. Set up anything that touches customer data without showing you exactly where it lives. How to stop: Stop booking. No exit interview. Skill stays with you. ## Page: about (https://scrappystart.ai/about) Scrappy Start is Erica and Jamie Sowder. Jamie does the teaching. He's spent twenty-plus years building software at Dell, eBay, and PayPal, and these days builds AI agents at CHAP Labs. A couple of those years he taught JavaScript at Hack Reactor, where he learned to teach a craft, not lecture about one. Erica runs outreach, community, and operations. She ran programs at UT Austin for ten years, then trained as a software engineer and worked at Indeed on their GenAI infrastructure. Two kids, two cats with cereal-themed names. They built this slowly, on purpose. How we work, in six lines: 1. Same shape, every hour. Sixty to ninety minutes. Both screens shared. We work on your real stuff. 2. You own everything. Prompts, setups, moves. All on your Claude account. 3. Stop by stopping. No retention call. 4. Plain language only. If we use a word you wouldn't use at the counter, we've failed. 5. One thing per session. The skill compounds when we don't try to fix everything at once. 6. Tools you already have. We build on top of your Claude, your inbox, your phone. ## Page: who this is for (https://scrappystart.ai/who-this-is-for) This is for the wedding photographer who's answering inquiries at 10pm. The planner coordinating eight vendors for a Saturday event. The florist whose inbox fills up during engagement season. The DJ who sends the same three questions to every couple. The day-of coordinator whose timeline lives in a Google Doc that's never quite current. The hair and makeup artist juggling trial bookings and day-of call times for a full bridal party. Not for: venture-backed founders, agencies wanting to white-label, or someone shopping for the cheapest option. --- ## Lead magnet: 5 AI Workflows That Save a Small Business 5 Hours This Week Free 18-page PDF available at https://scrappystart.ai/5-workflows.pdf (no login required, no email gate to read it). The five workflows in the PDF map directly to five of the agent-installable skills below: friday-inbox-triage (workflow 1), ig-captions-in-your-voice (workflow 2), sunday-brain-dump-to-monday-brief (workflow 3), vendor-reply-batch (workflow 4), one-star-review-sanity-check (workflow 5). The full prose, setup, and example outputs live in the skills. --- ## Scrappy Recipes (long-form, replace SaaS) Recipes are long-form, hand-authored pages that name a specific small-business pain and walk through the Claude move that replaces it. Each shipped recipe is built around one or more atomic Skills (listed below). The full prose for each recipe lives on its human page; the atomic prompt is published as a Skill that an agent can install on its own. ### Recipe No. 01: AI receptionist for the calls you're missing URL: https://scrappystart.ai/recipes/ai-receptionist Replaces: $29–$285 / mo AI receptionist SaaS Use instead: Claude.ai + a triage Skill For: Any small business whose phone is the front door. Plumbers, salons, medical practices, contractors, photographers, bookkeepers. Skills used: https://scrappystart.ai/skills/voicemail-triage.md ### Recipe No. 02: The bid that's been sitting since Tuesday URL: https://scrappystart.ai/recipes/bid-follow-up Replaces: $49–$369 / mo proposal & estimating SaaS Use instead: Claude.ai Project + your past winning bids For: Small remodelers, plumbing / HVAC / electrical service shops, roofers, tree care. Anyone writing two to ten proposals a month. Skills used: (none yet) ### Recipe No. 03: Customer scheduling for service businesses URL: https://scrappystart.ai/recipes/lawn-care-scheduling Replaces: $29–$699 / mo scheduling SaaS Use instead: Claude.ai Project + a customer-list sheet + a Skill For: Any service business with a recurring customer book. Lawn care, window cleaning, snow removal, pool service, cleaners, pet sitters. Skills used: (none yet) ### Recipe No. 04: Client management after the HoneyBook price hike URL: https://scrappystart.ai/recipes/wedding-pro-post-honeybook Replaces: $29–$129 / mo HoneyBook and friends Use instead: Claude.ai + Google Drive + Dropbox Sign + Stripe For: Wedding planners, photographers, DJs, florists, and the rest of the event-vendor world leaving HoneyBook. Skills used: https://scrappystart.ai/skills/wedding-inquiry-response.md ### Recipe No. 05: Step up from Excel without paying $100 a month URL: https://scrappystart.ai/recipes/excel-to-claude-crm Replaces: $15–$100+ / seat / mo CRM SaaS Use instead: Claude Cowork + your existing Excel or Sheets For: Anyone with 20–50 customers and a spreadsheet that's getting away from them. Skills used: https://scrappystart.ai/skills/customer-followup-radar.md ### Recipe No. 06: Dubsado handles the workflow URL: https://scrappystart.ai/recipes/dubsado-inquiry-triage Replaces: 45 minutes of manual reply-writing per inquiry Use instead: Claude Project + your packages doc For: Wedding photographers, planners, florists, DJs, and videographers who use Dubsado but spend 15–30 minutes writing every inquiry reply from scratch. Skills used: (none yet) ### Recipe No. 07: The leads you quoted in February URL: https://scrappystart.ai/recipes/peak-season-followup Replaces: Cold leads and a follow-up folder full of drafts Use instead: A lead sheet + Claude For: Wedding vendors with unbooked leads from January through March who lose them during peak season because the follow-up never gets written. Skills used: (none yet) ### Recipe No. 08: Twenty hours of editing URL: https://scrappystart.ai/recipes/gallery-release-reminder Replaces: An hour of writing per gallery delivery Use instead: Claude + your post-wedding notes For: Wedding photographers who deliver galleries on Pixieset, Pic-Time, or similar and want the delivery email to feel as personal as the photos. Skills used: (none yet) ### Recipe No. 09: Eight vendors. Eight referral notes URL: https://scrappystart.ai/recipes/vendor-referral-note Replaces: 45 minutes of post-event note-writing Use instead: Claude + your post-event observations For: Wedding planners, photographers, and vendors who want to build vendor referral relationships but never find time to write the notes after an event. Skills used: (none yet) ### Recipe No. 10: The 6-month follow-up you keep forgetting to send URL: https://scrappystart.ai/recipes/post-wedding-upsell Replaces: Album and anniversary session revenue left on the table Use instead: Google Calendar reminder + Claude For: Wedding photographers who offer albums or anniversary sessions but rarely follow up because the couple fell out of their head six months after the wedding. Skills used: (none yet) Full recipe index: https://scrappystart.ai/recipes --- ## Agent-installable skills Each skill below is a standalone markdown file with frontmatter following the Anthropic Claude Skills convention. The same content is served raw at the URL listed at the top of each skill, for agents that prefer one fetch per skill. Where a skill backs a Scrappy Recipe, the recipe URL is named inline. --- ### Skill: claude-stack-glossary URL: https://scrappystart.ai/skills/claude-stack-glossary.md Description: The Claude ecosystem in plain English. Names, features, scheduling, models, pricing, and which product belongs where. Current as of May 2026. Pin this to any Claude project that writes about Anthropic products so the names come out right. Use when: Before writing any email, blog post, recipe, or proposal that mentions Claude products. Or when a customer asks "wait, what's the difference between Cowork and Code." Or any time you're not sure whether the recurring thing is called a Scheduled Task or a Routine. Saves: the embarrassing reply where you tell a customer to "set up a Routine in Cowork" when Routines don't live there License: CC0 # The Claude stack glossary The Anthropic product family has gotten dense. There's Claude.ai, Claude Desktop, Claude Cowork, Claude Code, plus Skills, Projects, Scheduled Tasks, Routines, `/loop`, MCP, the Agent SDK, and Managed Agents. Half of those names sound interchangeable. They are not. This is the cheat sheet that keeps you (and your customers) on the right one. Last verified against primary Anthropic docs on May 13, 2026. If something below is wrong, [tell me](mailto:jamie@scrappystart.ai) and I'll fix it. ## The four surfaces There are four top-level places to run Claude. They all share your account login but they are not the same product, and a Skill you upload to one does not automatically show up in another. ### Claude.ai The default. A chat product on the web, iOS, and Android. Includes Projects (chat workspaces with custom instructions and files), Artifacts, Research, web search, Memory, and connectors. The model you talk to by default is Sonnet 4.6. This is the surface you use from your phone. It is also where most small business owners spend ninety percent of their Claude time. ### Claude Desktop The Mac and Windows native app. Looks like Claude.ai with extra hosting hooks. Drag a file in. Drop in a screenshot. Add local MCP servers via `claude_desktop_config.json`. The reason Desktop matters: **Cowork only runs inside Claude Desktop.** Not on the web. Not on the phone. If you want Claude doing real autonomous work for you, you need the desktop app installed on a Mac or Windows machine that is awake when the work runs. ### Claude Cowork Anthropic's product for non-coding agentic work. Lives **as a mode inside the Claude Desktop app**, not as a separate app. You switch into Cowork from a tab. It does file org, data extraction from screenshots and receipts, document drafting, light browser automation, Slack and Chrome integration, computer use as a fallback. Cowork is the SMB-friendly side of the agentic stack. The dev-shop side is Claude Code. Available on Pro, Max, Team, Enterprise. Not on Free. ### Claude Code The CLI and IDE tool that started as a coding assistant and is now the broader agent runtime. Lives in your terminal, on `claude.ai/code` on the web, and as IDE plugins (VS Code, JetBrains). Despite the name, Claude Code is not just for software developers. It's the right tool any time you want Claude to read your local filesystem, run scripts, schedule cloud-hosted automations (those are Routines, see below), or do an agentic task that doesn't fit the Cowork box. Included in Pro ($20). You do not pay extra for Claude Code on top of your Claude subscription. This is the single most-missed fact in the entire pricing page. ## Features that live across surfaces ### Skills (also called Agent Skills) A markdown file with frontmatter (`name`, `description`) plus optional supporting code and reference files. Claude loads the Skill on demand when a task matches the description. You can pin a Skill to a Project so it's always loaded in that workspace. Skills are account-level: upload a Skill once and it is available in Claude.ai, Cowork, Claude Code, and Office add-ins without re-uploading. You can also pin a Skill to a specific Project to make sure it always loads in that workspace. Anthropic ships pre-built Skills for `pptx`, `xlsx`, `docx`, and `pdf` document handling. You can also write your own, like the Scrappy Recipes on this site. ### Projects A claude.ai container that bundles custom instructions, knowledge files, and conversation history into one persistent workspace. Per-conversation context. Projects and Skills solve different problems. A Project is a workspace with your stuff in it. A Skill is a reusable instruction package the model loads when relevant. You can pin Skills to Projects. ### Custom Instructions Plain free-text personalization. Lives inside a Project, or at the account level. The lightest possible Claude customization. The mental model: **Custom Instructions** is a sticky note. A **Project** is a desk with that sticky note plus your files plus your past conversations. A **Skill** is a reusable manual you can put on any desk. ### MCP (Model Context Protocol) An open spec for plugging AI models into outside tools and data. Gmail. Google Calendar. Your file system. Custom apps. Slack. Anything. Supported on Claude.ai, Claude Desktop, Claude Code, and Cowork. The plain-English version: MCP is the connector layer that turns Claude from "a model that answers questions" into "an assistant that can actually touch your stuff." For small business owners, the practical entry point is Google Calendar and Gmail connectors. ## The scheduling muddle (the most confusing area, get this right) Three names. Three surfaces. Not interchangeable. | Name | Lives in | What it is | |---|---|---| | **Scheduled Tasks** | Cowork (inside Claude Desktop) | Recurring tasks that run on your Mac or PC. Requires the machine to be awake and Claude Desktop to be open at trigger time. Frequency: hourly, daily, weekly, weekdays. | | **Routines** | Claude Code (cloud-hosted) | Cloud-run automations. Trigger by schedule (≥1 hour cadence), API call, or webhook. Runs on Anthropic infrastructure. Your machine can be off. Per-day limits: 5 on Pro, 15 on Max, 25 on Team and Enterprise. | | **`/loop`** | Claude Code (a session command) | A polling command inside a single Claude Code session. `/loop 5m check the deploy` reruns the prompt every five minutes for the life of the session. Auto-expires after 7 days. Session-scoped, not persistent. | **Do not say "Routine" when you mean a Cowork Scheduled Task.** They are different products. The shorthand for "a thing on a clock" is "Scheduled Task" if you're in Cowork (most SMB scenarios) and "Routine" if you're in Claude Code (more technical, cloud-hosted). There is also a fourth thing called **Desktop Scheduled Tasks** that lives inside Claude Code as a local-runner variant. Runs on your machine on schedule but persists across restarts, unlike `/loop`. This one is a power-user feature; most people don't need to know it exists. ## Models, briefly | Model | What it's good for | |---|---| | **Claude Opus 4.7** | The flagship. Hardest reasoning, most expensive, default model on Enterprise. Available on Pro and up with the right model selector. | | **Claude Opus 4.6** | Previous flagship. Still available, still excellent, slightly cheaper. | | **Claude Sonnet 4.6** | The everyday workhorse. The default model on Claude.ai and Cowork. What you're talking to when you open the app and don't change anything. Released Feb 17, 2026. | | **Claude Haiku** | The smallest and fastest. For high-volume cheap-and-fast tasks. Not what you want for nuanced writing. | The 1M context window is available on Opus 4.6, Opus 4.7, and Sonnet 4.6 via the API and Claude Code cloud sessions. On Claude.ai chat the practical context cap is 500K for all paid plans, including Max. **Practical rule for SMB owners:** stay on Sonnet 4.6 (the default) for everything until you hit a wall. If a long document keeps getting cut off or a hard reasoning task feels weak, switch to Opus. ## Pricing, May 2026 | Plan | Price | What you get | |---|---|---| | **Free** | $0 | Basic Claude.ai chat, pre-built Skills (no custom upload), no Cowork, no Claude Code. | | **Pro** | $20/mo ($17/mo annual) | Full Claude.ai, **Cowork**, **Claude Code**, custom Skills. Routines: 5/day. The cheapest way to access the full toolkit. | | **Max 5x** | $100/mo | Everything in Pro plus higher rate limits, Opus 4.7 at 1M context, Routines: 15/day. The "I rely on Claude every day" tier. | | **Max 20x** | $200/mo | Same toolkit as Max 5x, higher rate limits again. For very heavy users. | | **Team** | $25/seat/mo ($20 annual). Premium seats $100/seat/mo annual ($125/seat/mo monthly). | Org features, admin controls, shared Skills via Skills directory. | | **Enterprise** | Custom, 50-seat minimum | Everything plus SSO, audit logs, default Opus 4.7. | **The fact most SMB owners miss:** Pro at $20 already includes Cowork and Claude Code. You don't need to upgrade for "the real tools." You need to learn what's already in your subscription. ## Common confusions, with the truth | You might think... | The truth | |---|---| | Cowork is a separate app | It's a mode inside the Claude Desktop app for Mac or Windows. Not on web or mobile. | | Claude Code is just for developers | No. It's any agentic workflow that needs filesystem access, scripted automation, or cloud Routines. SMB owners with one terminal open can use it. | | Skills sync across my Claude account | They do. Skills are account-level and available in Claude.ai, Cowork, Claude Code, and Office add-ins without re-uploading. | | Routine and Scheduled Task are interchangeable | They are not. Scheduled Task is Cowork. Routine is Claude Code cloud. | | I need Max to get Cowork | No. Pro ($20) includes Cowork and Claude Code. | | Claude Code SDK is the developer tool | It was renamed. The current name is **Claude Agent SDK**. | | Anthropic Console / Claude Console | The developer console is now at **platform.claude.com**, not console.anthropic.com. | ## Which surface for which job (SMB version) | The work | Surface | Why | |---|---|---| | Triaging customer emails, drafting replies, captions, light writing | **Claude.ai (web or mobile)** | Lightest weight. Phone-accessible. No setup. | | Reading a contract or screenshot, anything with files | **Claude Desktop** | Native drag-drop, plus local file access. | | "Run this every Monday morning while I sleep" on your real files | **Cowork Scheduled Task** | Local-machine agent on a clock. Easiest to set up. | | "Run this every weekday at 6am from the cloud, machine can be off" | **Claude Code Routine** | Cloud-hosted, doesn't need your machine. Slightly more setup. | | One-off agentic job, e.g., "go through these 200 invoices and pull totals" | **Cowork** (handed a task) | Cowork's bread and butter. | | Complex workflow with shell + filesystem + scheduled cloud + GitHub | **Claude Code** | Has every knob. | | "I'd rather not figure this out solo" | **A Scrappy Hour with me** | One hour, on Google Meet, $0 first or $150. | ## Things new since January 2026 If you read this with old context in your head, here's what changed. - **Sonnet 4.6** (Feb 17, 2026) replaced Sonnet 4.5 as the default Claude.ai model. - **Opus 4.7** (April 16, 2026) is the new flagship; Opus 4.6 still ships. - **1M context** went GA across Opus 4.6 / 4.7 / Sonnet 4.6. - **Routines in Claude Code** launched April 14, 2026. Cloud-hosted scheduled automations. Different product from Cowork Scheduled Tasks. - **Claude Code SDK was renamed Claude Agent SDK.** - **Managed Agents** beta launched (hosted REST API for production agents). - **Anthropic Labs** launched as an experimental product incubator. First public release: **Claude Design**. - **Cowork enterprise admin controls** launched April 9, 2026. - Help center moved from `support.anthropic.com` to `support.claude.com`. The product blog is at `claude.com/blog`; `anthropic.com/news` continues operating for company news. Both are active. ## Receipts Primary sources for everything above, with last-updated dates where Anthropic showed them. - [Claude pricing](https://claude.com/pricing) - [Pro plan, what's included](https://support.claude.com/en/articles/8325606-what-is-the-pro-plan) (≈ Apr 2026) - [Max plan, both tiers](https://support.claude.com/en/articles/11049741-what-is-the-max-plan) (April 7, 2026) - [Cowork product page](https://claude.com/product/cowork) - [Get started with Cowork](https://support.claude.com/en/articles/13345190-get-started-with-claude-cowork) (≈ late April 2026) - [Schedule recurring tasks in Cowork](https://support.claude.com/en/articles/13854387-schedule-recurring-tasks-in-claude-cowork) (April 9, 2026) - [Introducing Routines in Claude Code](https://claude.com/blog/introducing-routines-in-claude-code) (April 14, 2026) - [Run prompts on a schedule (Claude Code docs)](https://code.claude.com/docs/en/scheduled-tasks) - [Introducing Agent Skills](https://claude.com/blog/skills) (Oct 16, 2025; updated Dec 18, 2025) - [Agent Skills overview](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview) - [Use Skills in Claude](https://support.claude.com/en/articles/12512180-use-skills-in-claude) - [Claude Opus 4.7 announcement](https://www.anthropic.com/news/claude-opus-4-7) (April 16, 2026) - [Claude Sonnet 4.6 announcement](https://www.anthropic.com/news/claude-sonnet-4-6) (Feb 17, 2026) - [1M context GA](https://claude.com/blog/1m-context-ga) - [Context windows on paid plans](https://support.claude.com/en/articles/8606394) (≈ May 12, 2026) - [Agent SDK overview](https://code.claude.com/docs/en/agent-sdk/overview) - [Managed Agents overview](https://platform.claude.com/docs/en/managed-agents/overview) - [Model Context Protocol](https://claude.com/blog/what-is-model-context-protocol) - [Anthropic Labs intro](https://www.anthropic.com/news/introducing-anthropic-labs) ## Where this came from I help small business owners use Claude every week, and I kept getting the names wrong myself. This is the doc I now pin to every Claude project that writes about Anthropic products. It's CC0. Install it as a Skill in your Claude account, paste it into a Project as custom instructions, or share it with the bookkeeper who keeps mixing up Routines and Scheduled Tasks. If a name shifts or a price changes, this page updates. The version number in the frontmatter goes up. The receipts above stay current. Jamie at Scrappy Start --- ### Skill: friday-inbox-triage URL: https://scrappystart.ai/skills/friday-inbox-triage.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/friday-inbox-triage.md Description: Sort a week of customer emails into P1/P2/P3 with canned 2-sentence replies for the routine ones, in your voice. Use when: Friday afternoon, when you have 20+ unread customer emails and three of them actually need you. Saves: about 45 minutes per week License: CC0 # Friday inbox triage You spent forty-five minutes Friday afternoon reading customer emails. Most of them needed a two-sentence reply you've already sent twenty times. Three actually needed you. This skill finds the three in thirty seconds and pre-writes the rest. ## The setup 1. Open Claude. Start a new conversation. Free tier works for the first try. 2. Copy this week's customer emails as plain text. Anonymize anything you want. 3. Paste the prompt below, then your emails. ## The prompt ``` Below are this week's customer emails. For each one, tell me what the customer wants, whether it needs me, the priority, and the canned reply if applicable. My business: [one sentence: who I am, what I sell, where] My voice: [direct, friendly, no jargon, or your version]. I sign off with my first name, [Your Name]. PRIORITY rubric: - P1: customer is upset, refund/complaint in play, time-sensitive within 24 hours. NEEDS YOU is yes. REPLY is "(needs you)". - P2: real-business conversation, opens a revenue thread, needs me this week (wholesale, custom planning, partnership). - P3: routine recurring question. The 2-sentence canned reply covers it. Hours, shipping, returns, standard pricing, availability. Most emails are P3. When in doubt between P2 and P3, choose P3. Routine pricing and availability questions are P3 unless the customer signals urgency. OUTPUT FORMAT, same for every email: Email N. , . WANTS: NEEDS YOU: PRIORITY: REPLY: After the last email, end with: TALLY: P1, P2, P3. Do NOT invent ANY specific numbers in the REPLY (prices, package counts, turnaround times, deposits, minimums, distances). If a number isn't in my voice reference, write "[my rate]" or "[my standard window]" so I can fill it in. Better a placeholder than a made-up number a customer might quote back at me. NO em-dashes anywhere. No markdown bold, no asterisks, no preamble, no postamble, no commentary. Just the blocks and the tally. (Optional voice reference: paste 1-2 of your past replies here.) Emails: [paste them here] ``` ## What you'll get back For each email, a tiny block like this: > Email 1. Jordan M., asking what the travel fee is for a venue upstate. > Wants: pricing info on travel beyond the local area. > Needs you: no, canned reply works. > Reply: "Hi Jordan, travel within 60 miles is included. Beyond that I charge [my travel rate] per mile. Happy to give you an exact number once I know the venue." > Priority: P3. You read the three P1s. The other twenty have replies waiting to send. ## Adapt it to you Add one sentence to the prompt about your voice, like "warm, mentions customers by first name." Or paste two of your past replies and tell Claude to match the tone. The voice reference is the trick. Don't overthink the description. What you'd say at the counter is what you tell Claude. ## Where this came from Workflow 1 in the free [5-workflows PDF](https://scrappystart.ai/5-workflows). Visual version on Instagram: [the 9pm Friday inbox](https://www.instagram.com/p/DYC8gfekSXy). Screen-share Reel: [9pm Friday inbox, 45 minutes deep](https://www.instagram.com/p/DYGAQlaB5k7). Jamie at Scrappy Start --- ### Skill: voicemail-triage URL: https://scrappystart.ai/skills/voicemail-triage.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/voicemail-triage.md Description: Sort a voicemail transcript or missed-call text into one of five urgency buckets and draft the text reply you'd send back, in your voice, ready to paste into iMessage. Use when: When a new inquiry voicemail or missed-call text comes in while you're mid-shoot, in a venue walkthrough, or with a couple, and you need to triage and respond fast enough not to lose the lead. Saves: about 5 minutes per missed-call response, and the lead itself when urgency is right License: CC0 # Voicemail triage You miss a third of your calls when you're busy. The leads go to the next one on the list unless they hear back fast. This skill reads a voicemail transcript (or a missed-call text) and gives you back the urgency, the facts, and the exact reply to paste straight into iMessage. ## The setup 1. Open Claude. Start a new conversation. Free tier works for the first try. 2. Paste the prompt below into the message field, edit the bracketed placeholders for your business once, then paste the voicemail transcript (or the missed-call text) underneath. 3. Save the prompt as a Claude Project, a Cowork project, or a `SKILL.md` so you don't re-paste it every time. ## The prompt ``` You are a voicemail triage assistant for [BUSINESS NAME], a [BUSINESS TYPE: plumber, salon, medical office, photographer, contractor, bookkeeper, etc.] serving [SERVICE AREA]. Your job is to help [OWNER NAME] respond fast to missed calls without dropping leads. When the owner pastes in a voicemail transcript, a missed-call text from a customer, or a forwarded voicemail email, do three things in order. 1. Classify urgency into ONE of these five buckets. EMERGENCY: real-time damage, real-time harm, imminent risk of major damage (the next hour could turn bad), or a deadline that cannot wait. When in doubt between EMERGENCY and SAME-DAY, choose EMERGENCY. [EDIT FOR YOUR TRADE. Examples: plumber: burst pipe, leaking water heater with popping sounds, no heat in winter, gas smell, sewage backup; medical office: acute pain, post-op concern with infection signs, medication reaction with breathing trouble; salon: wedding-morning no-show, allergic reaction to a service; photographer: wedding-day equipment failure, no-show second shooter; contractor: roof leak in a storm, electrical sparking.] SAME-DAY: it stopped working today, customer is anxious but not in active or imminent damage. Needs a callback today. NORMAL: estimate request, scheduling a known job, general question. Callback by next business day is fine. UNCLEAR: the transcript is garbled, missing critical context, or you genuinely cannot tell. Do not guess. Ask the caller for more info. SPAM: robocall, telemarketer, wrong number, or anything that is not a real customer message. 2. Extract what you can find. - caller_name (if stated) - caller_phone (from the transcript metadata or the text sender if visible) - address or service area - one-sentence description of the problem - the time the message came in 3. Draft a text reply [OWNER NAME] can copy and paste straight into iMessage. Voice: warm, professional, brief. No emojis. NO EM-DASHES (use periods, commas, semicolons, or parentheses; em-dashes are an AI tell that customers notice). Use [OWNER NAME]'s first name. Per urgency: - EMERGENCY: confirm receipt, give a ten-minute callback window, ask for the address if not stated, ask them to text any photos if helpful. - SAME-DAY: confirm receipt, give a same-day callback window, ask for the address if missing. - NORMAL: confirm receipt, give a next-business-day callback window, no urgency. - UNCLEAR: ask the caller to text back with their address and a one-line description of the problem so [OWNER NAME] can call them back fast. - SPAM: leave the DRAFT TEXT REPLY line empty. Write "(none)" on its own line. Output format. Begin your response directly with the literal text "URGENCY:". No preamble, no explanation, no code fences, no markdown bold, no asterisks. Plain text only. Five labeled lines, in this exact order: URGENCY: SUMMARY: DRAFT TEXT REPLY: NEXT STEP FOR YOU: ``` ## What you'll get back For a transcript that reads "Hi, this is Alex, I found you on Instagram. My wedding is this Saturday and my photographer just canceled this morning. I'm kind of panicking. If you have any availability please call me back. My number is 312-555-7291," you'd get back something like: > URGENCY: EMERGENCY > SUMMARY: Alex (312-555-7291), Saturday wedding, photographer canceled same morning, needs last-minute coverage. > DRAFT TEXT REPLY: > Hi Alex, this is Jamie. I got your message. I'm looking at Saturday right now. Can you text me the venue, ceremony time, and how many hours you need covered? I'll get back to you in ten minutes. > NEXT STEP FOR YOU: > call back within ten minutes, confirm venue and hours before committing You read one line. You paste the reply. The couple hears back inside fifteen minutes instead of booking someone else. ## Adapt it The emergency examples in the prompt are the part you want to edit for your trade. Replace the bracketed list with your own four or five real emergencies, in your own language. After a month of running this on real voicemails, edit it once more for the two or three things Claude kept getting wrong. ## Where this came from Comes out of [Recipe No. 01, the AI receptionist for the calls you're missing](https://scrappystart.ai/recipes/ai-receptionist). The recipe is the long version with the deployment paths (a Claude Project on your phone, a scheduled morning batch, or a guided setup hour) and the SaaS context. This page is the standalone, portable prompt. Jamie at Scrappy Start --- ### Skill: repeat-question-replies URL: https://scrappystart.ai/skills/repeat-question-replies.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/repeat-question-replies.md Description: Generate canned replies in your own voice for the 3-5 questions you answer over and over every week. Use when: When you realize you're typing the same answers to "are you open Sunday" or "do you ship to Canada" five times a week. Saves: 2-3 minutes per repeat reply, plus the mental load License: CC0 # Repeat-question replies Most owners have three to five questions they answer over and over, in nearly the same words, every single week. The cost is forty-five seconds a reply, multiplied by every reply. This skill writes the canned versions in your voice once, so you can paste them forever. ## The setup 1. Open Claude. Start a new conversation. 2. Have ready: one sentence about your business, your three to five most-repeated questions, and one or two past replies that sound like you. 3. Paste the prompt below. ## The prompt ``` My business: [one sentence, e.g. "I photograph weddings in Chicago"] Below are the [3-5] questions I answer constantly. Write a canned reply for each one, in my voice. Warm, plain, no jargon. Sign off with my first name. OUTPUT FORMAT. For each question, output exactly: Q: A: Number them in order. No markdown bold around Q or A. No preamble ("Here are your replies..."), no postamble ("Let me know if you want changes..."), no commentary, no recommendations. Output ONLY the Q/A blocks, then stop. NO em-dashes anywhere in the replies. Em-dashes are an AI tell that customers notice. Use periods, commas, semicolons, or parentheses. Questions: 1. [question] 2. [question] 3. [question] Voice reference (1-2 of my past replies): [paste them] ``` ## What you'll get back Short, send-ready answers like: > Q: "Do you do engagement sessions?" > A: "Hi [Name], yes. A two-hour session runs [my rate] and usually happens a few months before the wedding. Most couples use it to get comfortable in front of the camera. Jamie" > > Q: "What's your travel fee for venues outside Chicago?" > A: "Hi [Name], travel within 50 miles is included. Beyond that I charge [my rate] per mile. Tell me the venue and I'll give you an exact number. Jamie" Five answers in your voice, in thirty seconds. ## Where to keep them so you actually use them Save the answers as text replacement on your phone. iPhone: Settings → General → Keyboard → Text Replacement. Type "ship" and the full canned reply lands. Friction goes from forty-five seconds to two. Or paste them into Gmail Templates, Apple Notes shortcuts, your CRM canned-reply field. Wherever your fingers go first. ## Where this came from Carousel #003 on Instagram: [the same three questions, every week](https://www.instagram.com/p/DYDK82Yj5VS). Jamie at Scrappy Start --- ### Skill: ig-captions-in-your-voice URL: https://scrappystart.ai/skills/ig-captions-in-your-voice.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/ig-captions-in-your-voice.md Description: Generate five Instagram caption drafts in your own voice from a few past captions and one sentence about what's happening this week. Use when: When you're staring at the IG caption box at 9am Monday with a perfectly good photo and nothing to write. Saves: about 20 minutes per post License: CC0 # IG captions in your own voice You stared at the IG caption box for twenty minutes. Wrote three lines, deleted them, posted nothing. Or posted something that didn't sound like you. This skill drafts five options in your voice in thirty seconds. You stay the editor. ## The setup 1. Find two or three of your past captions that got real engagement. Copy them. 2. Open Claude. Start a new conversation. 3. Have one sentence ready about what's happening this week (new bread, an open chair Friday, a print job shipping). ## The prompt ``` Here are 3 captions of mine that worked well for my [wedding photography / planning business / floral studio / DJ company]: [paste your 3 captions] The voice: [one sentence: warm, specific, mentions couples by name, never asks for likes] This week: [what's happening, e.g. "delivering the first spring gallery of the season" or "fall 2027 booking just opened"] Write 5 caption options in that voice. Vary length and angle across the 5 so I have real choice (one shorter, one with a story moment, one straightforward). USE ONLY THE FACTS in the "This week" line above. Do not invent the topic of a meeting, the type of training, a forthcoming product, an event detail, a customer name, or anything not present in my input. If a detail is missing, write around it. OUTPUT FORMAT. Plain text only, no markdown bold, no preamble, no postamble, no recommendation on which option to pick: OPTION 1: OPTION 2: OPTION 3: OPTION 4: OPTION 5: No emojis. No "link in bio" CTAs. No em-dashes (em-dashes are an AI tell that followers notice; use periods, commas, parentheses). ``` ## What you'll get back Five drafts, each ready to post or edit lightly. For a photographer delivering the first spring gallery of the season, one draft might be: > Six months of weddings in a folder. Yesterday was delivery day. Cherry blossoms at 7am. First dance at 10pm. The Riveras got their gallery this morning. Pick the one closest to you. Edit one line. Post. ## To make it sound even more like you Paste the closest draft back to Claude. Ask: "Shorten by 20% and use the word [a word you actually say a lot]." Claude refines. The more of your voice in the prompt, the less you edit at the end. ## Where this came from Workflow 2 in the free [5-workflows PDF](https://scrappystart.ai/5-workflows). Visual version on Instagram: [the caption you didn't write](https://www.instagram.com/p/DYDL5GCDyf1). Jamie at Scrappy Start --- ### Skill: sunday-brain-dump-to-monday-brief URL: https://scrappystart.ai/skills/sunday-brain-dump-to-monday-brief.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/sunday-brain-dump-to-monday-brief.md Description: Take a 5-10 minute stream-of-consciousness dump of everything in your head about the week ahead and get back a 1-page Monday-morning brief with must-dos, what can wait, and a 15-minute first step on the thing you're procrastinating. Use when: Sunday at 8pm, when you're already dreading Monday and carrying everything-not-yet-decided into bed. Saves: 2-3 hours of avoidance per week, plus the dread License: CC0 # Sunday brain dump to Monday brief It's Sunday at 8pm. You're already dreading Monday. The problem isn't the work, it's carrying everything-not-yet-decided into bed. This skill empties your head onto the page and sends back a one-page brief you didn't have to write. ## The setup 1. Open Claude. Go to a quiet room. 2. Brain-dump for five to ten minutes. Type it, or use Claude's voice mode (the soundwave icon next to the microphone; the microphone alone is plain dictation, voice mode is the back-and-forth conversational one) and just talk. Voice mode is faster and matches the stream-of-consciousness energy better than typing. Don't edit. Bookings, quotes, the thing you keep meaning to call about, the vendor that owes you, the email you've been avoiding. 3. Run the prompt below over your dump. ## The prompt ``` Sort the following brain dump into four sections. OUTPUT FORMAT. Use these exact section headers, in this order, plain text only, no markdown bold, no preamble, no postamble, no commentary: MUST DO THIS WEEK <3 to 5 items max. Each item is a concrete verb plus specifics (not "follow up on the kitchen" but "send the Mendoza kitchen quote with the revised line items"). One per line.> CAN SHIFT TO NEXT WEEK WORRIED ABOUT, AND WHY MONDAY'S FIRST MOVE End with one line: "Screenshot this and put it where you'll see it Monday morning." Plain English. No em-dashes (use periods, commas, parentheses). No consultant words ("leverage," "synergy," "streamline," "optimize"). [your brain dump] ``` ## What you'll get back A clean brief, in your size. For a wedding photographer's Sunday-night dump, Claude might respond with: > Must do this week (4 items) > 1. Send the Huang wedding contract. They said yes last Friday. Don't let the slot drift. > 2. Email the second shooter about the Martinez setup time. Need a confirmed arrival window. > 3. Finish culling the Patel gallery before the 10-day delivery window closes. > 4. Reply to the three October inquiries sitting in the "pricing" folder. > > Can shift to next week > Portfolio update. New pricing page. The business card redesign. > > What you're worried about > The Huang contract quote is slightly lower than what the timeline actually requires. Send the contract with a line-item note on the hours so the scope is explicit before the deposit clears. > > The thing you keep procrastinating on > Following up on February leads. Monday's 15-minute step: open the inquiry spreadsheet, find the three couples who went quiet, send a one-paragraph check-in to each. Don't overthink it. Just send. ## Make it a Sunday ritual Pick a Sunday time you'll keep (after lunch, before kids' bedtime). Open Claude. Run the prompt. Screenshot the brief and stick it where you'll see it Monday morning. After two Sundays, the dread stops showing up uninvited. ## Where this came from Workflow 3 in the free [5-workflows PDF](https://scrappystart.ai/5-workflows). Visual version on Instagram: [the Sunday-night brain dump](https://www.instagram.com/p/DYDv7BZDAMb). Screen-share Reel: [Sunday at 8pm, already tired](https://www.instagram.com/p/DYI3PMvB1ZD). Jamie at Scrappy Start --- ### Skill: vendor-reply-batch URL: https://scrappystart.ai/skills/vendor-reply-batch.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/vendor-reply-batch.md Description: Write four short, send-ready vendor replies in two minutes by giving Claude one sentence per email about what you want the reply to do. Use when: When you owe four vendors a reply and you don't want to write four emails. Saves: about 30 minutes per week License: CC0 # Vendor reply batch You owe four vendors a reply. You don't want to write four emails. This skill turns one sentence per email into four short, send-ready drafts in your voice. ## The setup 1. Open Claude. Start a new conversation. 2. Forward or paste each vendor email into a single conversation. 3. After each one, write one sentence about what you want the reply to do. ## The prompt ``` Below are 4 vendor emails I owe a reply to. For each one, here's what I want my reply to do. Vendor 1 email: [paste] What I want the reply to do: [one sentence] Vendor 2 email: [paste] What I want the reply to do: [one sentence] (Continue for all 4.) Write the 4 replies. Keep them short. My voice is direct, friendly, no jargon. Sign off with my first name. OUTPUT FORMAT (plain text, no markdown bold, no preamble, no postamble, no commentary): REPLY 1: REPLY 2: (continue for REPLY 3 and REPLY 4) End with: "End of batch." CRITICAL. The output MUST be send-ready: - If the vendor's signature gives you a name, USE IT. If not, open with "Hi," (no name, no [bracket]). - NEVER leave bracketed placeholders ([name], [address], [FILL IN]) in the output. Brackets in the output are bugs. - NEVER write meta-asides to me ("if you provide the address I'll include it" or "let me know and I'll add X"). If a detail is missing, DROP the sentence that needed it, or rephrase without it. - If you genuinely cannot write a reply without missing context, output "(needs you)" for that reply and move on. NO em-dashes anywhere in the replies. Em-dashes are an AI tell that vendors notice. Use periods, commas, semicolons, or parentheses. ``` ## What you'll get back Four short, send-ready replies. One might read: > To: Vendor 2 (the one offering the new lease terms) > > Hi David, > > Thanks for sending the new terms over. The 18-month at the same rate works for us. Send the contract when you're ready and I'll get it back to you Monday. Appreciate you holding the rate. > > Jamie Read each one in five seconds, click send. About two minutes total. ## Where this came from Workflow 4 in the free [5-workflows PDF](https://scrappystart.ai/5-workflows). Jamie at Scrappy Start --- ### Skill: vendor-scope-creep-pushback URL: https://scrappystart.ai/skills/vendor-scope-creep-pushback.md Description: Reply to "while you're at it" requests that protects the original scope AND the relationship. Acknowledges the ask, names what's in and out, offers a concrete option, keeps the warmth. Use when: When a client or vendor adds a "small thing" to a job that wasn't in the original agreement, and you don't want to silently absorb it OR sound like a contract lawyer. Saves: hours per month, plus the mental load of saying yes when you should have said "yes for $X" License: CC0 # Vendor scope-creep pushback You agreed to photograph the ceremony and reception. Now they want getting-ready coverage added too, "since you'll already be there." Most owners feel two bad options: silently absorb (eat the cost), or sound like a contract lawyer (lose the client). There's a third reply, and you keep not writing it. This skill writes it. ## The setup 1. Open Claude. Start a new conversation. 2. Compress the original agreement and the new ask into one sentence each. Don't paste the whole email thread. Two sentences is the trick. 3. Decide what your "yes for $X" option is, or whether you'd rather refer them out. You can also let Claude propose both. ## The prompt ``` I'm a [wedding photographer / wedding planner / florist / DJ / videographer / day-of coordinator]. A [client / vendor] asked me to do something that wasn't in our original agreement. The original agreement (one sentence): [e.g. "photograph the ceremony and reception, 8 hours, $3,200"] What they're now asking for (one sentence): [e.g. "also cover getting-ready, 'since you'll already be there early'"] Write me a short reply that: 1. Acknowledges the ask warmly (no defensiveness) 2. Names what's in scope and what isn't, in plain words, no contract-speak 3. Offers a concrete option (a quote for the new work, a referral to someone who handles it, or "yes but for $X") 4. Keeps the relationship warm My voice is direct, friendly, signs off with my first name. Don't apologize for the scope. The scope is the scope. ``` ## What you'll get back A short reply, ready to send. For the getting-ready add-on example: > Hey Jordan, glad we're locked in for the day. Getting-ready coverage isn't in the original package, but I can add it. It's typically two hours and I'd invoice it at $600. I could start at the hotel at 10am instead of the ceremony at noon. Want me to add it to the contract? If you'd rather keep the original scope, we're completely set as-is. > > Jamie Same warm tone. Different number. Cleaner week. ## Adapt it for your trade The shape works for: subcontractors, freelancers, studio rentals, design clients, custom orders, wedding photographers, accountants. Save the prompt with the words for your work. "While you're at it" looks different in every trade. The reply has the same shape every time. ## Where this came from Comes up in nearly every Scrappy Hour with anyone who quotes work. The skill complements [vendor-reply-batch](https://scrappystart.ai/skills/vendor-reply-batch), which is about reply VOLUME. This one is about reply BOUNDARY. Visual version on Instagram: [the "while you're at it"](https://www.instagram.com/p/DYFGywBEV5G). Jamie at Scrappy Start --- ### Skill: wedding-inquiry-response URL: https://scrappystart.ai/skills/wedding-inquiry-response.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/wedding-inquiry-response.md Description: For wedding planners, photographers, DJs, florists, and event vendors. Takes a new inquiry (contact form, IG DM, email) and drafts a warm personal first reply with the right discovery questions, the package that fits, and a suggested next step. In your voice. No template-email energy. Use when: When a wedding or event inquiry lands and you need to send a real-feeling reply within an hour, not three days from now. Saves: about 20 minutes per inquiry, plus the bookings you would have lost to whoever responded faster License: CC0 # Wedding inquiry response Wedding photographers and planners get a lot of inquiries. The ones who book are usually the ones who get a warm, specific reply inside an hour, not a templated form letter from HoneyBook three days later. This skill takes the inquiry, drafts the reply in your voice, asks the right one or two questions to keep the conversation moving, and suggests the next step. You read it, fix one line, send. ## The setup 1. Open Claude. Start a new conversation. Free tier works for the first try. 2. Copy the inquiry as plain text. Wherever it came from: the contact form on your site, an IG or Facebook DM, a referral email. 3. Have one sentence ready about each of your packages with rough price ranges (or however you describe price publicly). ## The prompt ``` I'm a [wedding photographer / wedding planner / florist / DJ / videographer / day-of coordinator] based in [SERVICE AREA]. My voice is [warm, specific, mentions couples by name; or your version]. I sign off with my first name, [Your Name]. My packages (one line each, your price phrasing): 1. [e.g. "Six-hour wedding coverage: $3,200, 400 retouched images, online gallery"] 2. [...] 3. [...] A new inquiry came in. Draft a warm personal first reply that: - Greets the inquirer by name if they signed off with one - Acknowledges one specific thing from their message (the date, the venue, the vibe, the season, what they said about each other) - Mentions the package that fits, with my actual phrasing - Asks one or two discovery questions I'd want answered before pricing for real (guest count, ceremony location vs reception location, total event window, any quirks I should know about) - Suggests a next step: a 20-minute video call, or my booking link, or whatever you decide fits the inquiry shape - Stays under 180 words Voice rules: - NO em-dashes anywhere (em-dashes are an AI tell that couples notice; use periods, commas, semicolons, or parentheses) - No emojis unless I asked for them above - No "Just following up" or "Hope this finds you well" - No exclamation points beyond at most one - No marketing-speak ("elevate your special day," "unforgettable experience," "perfect package"). Specific over abstract. Inquiry: [paste the inquiry here] ``` ## What you'll get back A reply you can fix one line on and send. For a photographer's inquiry "Hi! I'm looking for someone to photograph our small backyard wedding on Sept 12 in Hudson Valley. We're around 40 guests, super casual. Loved your work with Tess + Marc. -Lana," you might get back: > Hi Lana, > > Thanks for reaching out. A 40-guest backyard wedding in the Hudson Valley in September is exactly the kind of day I love photographing, and I'm so glad the Tess and Marc gallery hit. I'm currently available on September 12. > > The six-hour coverage package is probably the right fit: $3,200, 400 retouched images, online gallery. Before I quote firmly, two questions. What time is the ceremony and how long are you imagining the reception going? And is there a specific moment you want me there for the start of (first look, getting ready, etc.)? > > Happy to hop on a 20-minute video call this week if that's easier than going back and forth. Here's my booking link. > > Jamie Specific, warm, no fluff, asks the right things, suggests the next step. Three minutes to read and send. ## Adapt it - For inquiries where the date is already booked: same skill, but ask Claude to draft the kind "I'm not available that date, here's who I'd recommend" reply, signed off with the same warmth so the couple remembers you next time. - For inquiries that look like price-shoppers (no specifics, just "what do you charge"): tell Claude to keep the reply shorter and ask one clarifying question before quoting. - For inquiries from referrals (somebody you know mentioned them): tell Claude to acknowledge the referral by name. Two-degree relationships book at much higher rates than cold inquiries; honor them. ## Where this came from Central skill in the wedding-pro-post-HoneyBook recipe at [scrappystart.ai/recipes/wedding-pro-post-honeybook](https://scrappystart.ai/recipes/wedding-pro-post-honeybook). The recipe is the long version: replaces HoneyBook with Claude + Google Drive + Stripe + Dropbox Sign for the parts Claude can't do (payments, e-signature). This page is the standalone, portable prompt. Jamie at Scrappy Start --- ### Skill: wedding-package-quote URL: https://scrappystart.ai/skills/wedding-package-quote.md Description: For wedding photographers, planners, florists, and DJs. Takes an incoming inquiry plus two or three of your past proposals and drafts a tailored first-reply package quote for the couple, in your voice. Feels personal. Takes ten minutes, not an afternoon. Use when: When an inquiry has enough detail to quote (date, venue, rough guest count or vibe) and you want to send a package proposal that doesn't read like a template. Saves: 30 to 45 minutes of blank-page proposal writing per inquiry, plus the bookings you'd lose to whoever replied faster License: CC0 # Wedding package quote Most proposal tools give you a template and call it personalization. A template-energy reply to a warm inquiry is a booking left on the table. This skill takes the new inquiry plus two or three of your past winning proposals and drafts a quote that reads like you wrote it for this couple specifically. You read it, fix one line, send it. Works for photographers, planners, florists, DJs, and anyone else who sends a package or pricing proposal to a new inquiry. ## The setup 1. Open Claude (Claude.ai, Claude Desktop, or wherever you've installed this skill). Start a new chat. 2. Attach two or three past proposals you sent to booked couples. PDFs, Word docs, screenshots, plain text copied from email. Pick ones that resulted in bookings. They become the reference Claude reads when drafting. 3. Attach or paste the new inquiry as plain text. 4. Paste the prompt below, replace the bracketed placeholders once for your business, and save it as a Claude Project's Instructions or a `SKILL.md` so you don't re-paste it every time. ## The prompt ``` You are a proposal drafter for [BUSINESS NAME], a [BUSINESS TYPE: wedding photographer / planner / florist / DJ / videographer / day-of coordinator] based in [SERVICE AREA]. Your job is to help [OWNER NAME] turn a new couple inquiry into a warm, tailored first-reply package quote in ten minutes, by re-using the voice and structure from past proposals the owner has attached to this chat. When the owner attaches a new inquiry (email, contact form, IG DM, or pasted text), do five things in order. Two rules that apply to every step. - Do not invent specific dollar amounts. If the inquiry asks about pricing and the owner's past proposals show a range, describe the range as it appears in the past proposals. Never make up a number. Better a placeholder than a price that surprises the couple. - Do not invent a phone number, portfolio link, or contact detail. Use the literal placeholder [CONTACT_INFO] so the owner can paste their real details before sending. 1. Read every past proposal the owner has attached to this chat. Identify the one or two that match the new inquiry most closely (similar event type, similar scale, similar season or venue style). Note them by filename in your MATCHED PROPOSALS line. If nothing is a close match, say so plainly and ask the owner to attach a closer past proposal. 2. Extract the couple's facts from the inquiry as bullets. - Names (if signed or mentioned) - Event date (if given) - Venue (if mentioned) - Guest count (if mentioned) - Vibe, tone, or any detail they shared about the day - Which package or service they asked about (if any) - What hasn't been said but will matter (event duration, second locations, travel, add-ons the inquiry implies) 3. Identify the best package fit. Name the package from the owner's past proposals that best fits this couple's inquiry. If two packages could fit, name both with one sentence on the difference. Flag any PACKAGE GAPS (the couple mentioned something the owner's past proposals don't clearly cover). 4. Draft the reply. A warm, personal first reply in [OWNER NAME]'s voice. Under 200 words. Open by greeting them by name (or "Hi there" if unsigned). Acknowledge one specific detail from their message. Name the package that fits. Ask one or two real discovery questions (not generic ones). Suggest a concrete next step (a video call, a tour of the venue portfolio, a booking-link invite). Sign off with [OWNER NAME]'s first name. Rules for the reply: - No em-dashes (they read as AI-generated; use periods, commas, or parentheses instead) - No "Just following up," "Hope this finds you well," "unforgettable," "elevate," "perfect day," or similar - No emojis unless the owner's past replies use them - No exclamation points - Zero invented prices or dates 5. List two or three questions [OWNER NAME] should ask this couple before the reply becomes a firm proposal. These should be the gaps you noticed while reading the inquiry and comparing to past proposals. Concrete and answerable, not vague. Output format. Begin your response directly with the literal text "MATCHED PROPOSALS:". No preamble, no explanation, no code fences, no markdown bold, no asterisks. Plain text only. Five labeled sections, in this exact order: MATCHED PROPOSALS: COUPLE FACTS: PACKAGE FIT: DRAFT REPLY: QUESTIONS BEFORE SENDING: <2 to 3 questions to confirm before the proposal is firm> ``` ## What you'll get back Five sections in plain text. The most useful is usually the QUESTIONS BEFORE SENDING, because they surface what the couple didn't tell you that changes the package. COUPLE FACTS is the sanity check. PACKAGE FIT tells you if you're reaching for the right package before you quote it. DRAFT REPLY is the actual email to send. For a photographer getting a short inquiry from a couple getting married in September at a rooftop venue: the draft might read: > Hi Jordan and Alex, > > The rooftop is a great pick for a September evening. I love shooting that kind of ambient light. > > Based on your date and what you mentioned, the Eight-Hour Coverage package is probably the right fit. It covers ceremony through first dances and gives you a second location if you want to do portraits somewhere downtown first. > > A couple of things worth knowing before I send the full quote: will you need a second shooter, and are you planning any portraits at a separate location before the ceremony? > > Happy to hop on a quick call this week if that's easier. Here's my calendar link: [CONTACT_INFO] > > Sarah Warm, specific, signed. You read it, fix one line if something's off, send it. ## Adapt it The two things to tune after a real month of inquiries are the voice description and the package names. Both are bracketed in the prompt. Your actual package names matter more than descriptions of them. If you have three packages, paste their real names and a one-line description of each into the Project Instructions alongside this prompt. Claude will match the inquiry to the right one instead of guessing. If you want to skip the five-section format for simpler inquiries and just get the draft reply fast, tell Claude after pasting the inquiry: "skip the sections, just write the draft reply." ## Where this came from Comes out of the Dubsado inquiry triage recipe at [scrappystart.ai/recipes/dubsado-inquiry-triage](https://scrappystart.ai/recipes/dubsado-inquiry-triage). The recipe is the long version with the deployment paths (a Claude Project you set up once and open on your phone for every new inquiry) and the SaaS context. This page is the standalone, portable prompt. Jamie at Scrappy Start --- ### Skill: quote-followup-without-being-pushy URL: https://scrappystart.ai/skills/quote-followup-without-being-pushy.md Description: Three follow-up email options for a quote that's gone quiet. Soft check-in, decision-time nudge, and graceful close, all in your voice. Use when: When you sent a quote 8+ days ago and they've gone silent, and you need to follow up without seeming desperate or losing the lead. Saves: about 20 minutes per quote follow-up plus the mental tax of avoiding it License: CC0 # Quote follow-up without being pushy You sent a quote eight days ago. They've gone silent. You don't want to seem desperate. You also can't afford to ghost back, because the silence kills more deals than the no. This skill writes the three follow-up emails you've been putting off, each one calibrated to a different stage of the silence. ## The setup 1. Open Claude. Start a new conversation. 2. Pull up the original quote and their last message. Compress the quote into one sentence: scope, price, timeline. 3. Decide which one of the three (or all three) you'll save and use over and over. ## The prompt ``` I sent a [quote / proposal / estimate] to a prospective customer [N] days ago. They've gone quiet. The quote (one sentence): [e.g. "eight-hour wedding photography package, $4,200, gallery in six weeks, available on their date"] Their last message: [paste it, or "they replied 'sounds good, let me think about it' and went quiet"] Write me three short follow-up email options: 1. The "soft check-in" (8-10 days after quote, no pressure) 2. The "decision-time" nudge (3-4 weeks after quote, asks them to commit or release the slot on my calendar) 3. The "graceful close" (6+ weeks, lets them know I'm putting the slot to someone else and the door is still open later) My voice is direct, friendly, no jargon. Sign with my first name. Don't start any of them with "Just following up" or "Bumping this." Don't apologize for following up. ``` ## What you'll get back Three short emails ready to send. The first might read: > Subject: Quick check on the photography quote > > Hi Jordan, wanted to make sure the quote landed on your end. No rush on the decision, just don't want it to disappear into the inbox. Happy to jump on a quick call this week if useful. > > Jamie The third (graceful close) might read: > Subject: Releasing your date hold > > Hi Jordan, the September 14 date I'd held for you is starting to get inquiry requests from other couples, so I'm going to open it back up. If you decide to move forward on a different date, reach out and I'll check availability. Either way, hope the planning is going well. > > Jamie ## Adapt it The decision-time nudge is the one most owners dodge. It feels like pressure. It isn't. You're protecting your calendar AND giving them permission to commit. Save that one and use it monthly. ## Where this came from Comes up in nearly every Scrappy Hour with contractors, photographers, and freelancers. Pairs with [vendor-reply-batch](https://scrappystart.ai/skills/vendor-reply-batch) for the rest of the email flow. Jamie at Scrappy Start --- ### Skill: customer-followup-radar URL: https://scrappystart.ai/skills/customer-followup-radar.md Guided install (paste into a new Claude chat): https://scrappystart.ai/skills/install/customer-followup-radar.md Description: For any small-business owner with a customer list in a spreadsheet (Excel, Google Sheets, Numbers, or even a Notes doc). Reads your list, identifies the 5 to 10 customers most worth a personal followup this week, and drafts the outreach in your voice. Replaces the "who haven't I talked to in a while" mental tax that pushes owners to overpay for a CRM. Use when: Monday morning when you want to do one round of customer followup, or any time you're staring at the spreadsheet and don't know who to start with. Saves: about an hour of decision-fatigue per week, plus the customers you would have let go cold License: CC0 # Customer followup radar You have 20 to 50 customers in a spreadsheet. You know you should follow up with some of them, you can't remember who, and the list is too long to scan in your head. Most CRMs solve this with a dashboard you pay $100 a month for. Claude solves it with a chat that reads your spreadsheet and tells you the five to ten people worth reaching out to this week, with drafted outreach for each one. Keep using your spreadsheet; this skill just reads it. ## The setup 1. Have your customer list ready. One row per customer, columns: Name, Email or Phone, Last Contact Date, Customer Type (Prospect / Active / Lapsed), Value or Segment, Notes (optional). 2. Open Claude Desktop and switch to Cowork. Attach your spreadsheet to the conversation, or drag the file in. Or paste the list as plain text if it's short. 3. Paste the prompt below with one sentence about your business and your typical voice. ## The prompt ``` My business: [one sentence: who I am, what I sell, where, what kind of customers] My voice: [direct, friendly, no jargon; or your version]. I sign off with my first name, [Your Name]. My customer list is attached (or pasted below). One row per customer. When I ask "what should I do this week?" (or similar), do four things in order. 1. RADAR. Identify the 5 to 10 customers most worth reaching out to this week. Pick using these rules in order: - Active customers I haven't talked to in 60+ days - Lapsed customers I haven't tried to win back in 90+ days - Prospects who went quiet without a clear no - High-value customers whose cadence has slipped - Anyone with a note flagging a specific reason to circle back For each, give one sentence on WHY they're on the list. 2. PRIORITIZE. Order the radar list by which feels most worth my 30 seconds first (highest-value reactive customers and quietly- slipping high-value ones go to the top). 3. DRAFT OUTREACH. For each customer in the radar list, draft a short email or text (under 80 words) in my voice. Reference any note I had on them. Sign off with my first name. No emojis. 4. QUICK STATS. Three to five numbers I should know about the list this week: - Customers I haven't touched in 60+ days - Lapsed customers (no activity in 90+ days) - High-value customers whose cadence has slipped - Any other one-line observation worth flagging OUTPUT FORMAT. Plain text only. No markdown bold. No preamble, no postamble, no commentary. Use these exact section headers: RADAR DRAFT OUTREACH QUICK STATS <3 to 5 short observations as bullets> NO em-dashes ANYWHERE in your output. Not in RADAR, not in DRAFT OUTREACH, not in QUICK STATS. Use periods, commas, parentheses, or semicolons instead. This applies to every section, not just customer-facing text. ``` ## What you'll get back A focused weekly plan you can scroll on your phone in two minutes. RADAR up top so you can spot the five you should actually do today. DRAFT OUTREACH ready to copy-paste into your email or text app. QUICK STATS so you know whether the list itself is healthy. For a wedding photographer's contact list, the radar might look like: > 1. Chen Wedding (high-value Active; last contact 67 days ago, gallery delivered, album conversation never happened) > 2. Taylor and Morgan (Prospect; went quiet after March 20 inquiry, no response to quote) > 3. Reyes Anniversary (Lapsed; booked last May, haven't circled back for this year's anniversary portraits) > ... And the drafts might read: > Chen Wedding: > Hi Jess, I've been meaning to circle back since the gallery went out. I'd love to chat about an album before the year gets away. No rush on a decision, just want to make sure it's on your radar. Jamie Specific, warm, signed. You read each, fix one line if needed, paste and send. ## Adapt it - For a B2B list with deal-stage notes, add a column "Stage" and tell Claude to weight stage progression in the RADAR rules. - For a service business with recurring engagements, replace "60 days no contact" with your typical cadence (every three months for past wedding clients you want album or referral conversations with). - For a high-volume list (100+ customers), tell Claude to surface 10 to 15 per week instead of 5 to 10. ## Where to keep the spreadsheet The spreadsheet is the database. Claude is the chat that reads it. Keep the spreadsheet in a Drive folder or iCloud folder you back up; the chat will re-read it whenever you re-attach. The list is yours forever, exportable to any tool you might want later. The chat is the brain you can rebuild from a fresh paste of this prompt at any time. ## Where this came from This is the central skill in the Excel-to-Claude-CRM recipe at [scrappystart.ai/recipes/excel-to-claude-crm](https://scrappystart.ai/recipes/excel-to-claude-crm). The recipe is the long version with the cost-comparison against HubSpot, Pipedrive, Salesforce, and the smaller modern CRMs; this page is the standalone, portable prompt. Jamie at Scrappy Start --- ### Skill: annual-rate-increase-letter URL: https://scrappystart.ai/skills/annual-rate-increase-letter.md Description: A short letter to existing clients announcing a rate increase, plainly. No buried lede, no over-apologizing, no guilt. Acknowledges the relationship and closes warm. Use when: When you've decided to raise your rates and you're scared to send the email, especially to your long-time loyal clients. Saves: weeks of procrastination plus the revenue you've been leaving on the floor License: CC0 # Annual rate increase letter You've been charging the same rate for three years. Rent went up. Your materials cost more. You're working harder for less, and you know it. The new rate is fair. You've also been putting off the email for six months because you don't want to lose the loyal client. This skill writes the letter. ## The setup 1. Open Claude. Start a new conversation. 2. Decide on the new rate, the effective date (give clients at least four weeks), and the reason in plain English. Not jargon. Not "due to rising operational costs." The actual reason, the one you'd say at the counter. 3. Have one sentence about the relationship if you want to acknowledge it. ## The prompt ``` I'm a [wedding photographer / wedding planner / florist / DJ / videographer / day-of coordinator] and I'm raising my rates. Current rate: [e.g. "$3,200 / wedding coverage"] New rate: [e.g. "$3,800 / wedding coverage"] Effective date: [e.g. "January 1"] The reason in plain English: [one sentence the client would understand, e.g. "gear and editing costs have gone up, I've added a second shooter option, and my last raise was three years ago"] Optional: this is going to a long-time client I've worked with for [N] years. Write me a short letter: 1. Tells them the new rate, plainly, no buried lede 2. Says when it takes effect and gives them at least 4 weeks 3. Acknowledges the relationship if it's a long-time client 4. Doesn't apologize. Doesn't over-explain. Doesn't guilt them. 5. Closes with "looking forward to seeing you next month" energy, not "thank you for your understanding" energy My voice is warm, direct, no jargon. Sign with my first name. ``` ## What you'll get back A clean, send-ready letter. For a photographer raising wedding coverage from $3,200 to $3,800: > Hey Rachel, > > Quick note about a change starting January 1. After three years at the same rate, I'm moving wedding coverage from $3,200 to $3,800. The short reason: gear costs have gone up, editing time is longer than it used to be, and it's been three years. > > You've sent me two referrals in the past year and I wanted to give you the heads up directly before I update the site. Same shooting style, same editing, same turnaround. Just a different number. > > Looking forward to the spring season. > > Jamie ## Adapt it Two things owners get wrong: burying the new number ("starting in the new year, there will be some changes to my pricing structure." Just say the number), and apologizing ("I'm so sorry to have to do this." You're not sorry, you're running a business). The letter above does neither. Save the prompt and use it every time you raise. ## A note on timing Send rate-increase letters at least four weeks before the effective date. Personal clients (stylists, therapists, coaches) deserve six. Don't bury the announcement at the bottom of a holiday newsletter. Treat the letter like a real piece of communication and most clients will respect it. The ones who leave were going to leave anyway. ## Where this came from The most-procrastinated email in small business. Comes up in nearly every Scrappy Hour with service providers eventually. Pairs with [quote-followup-without-being-pushy](https://scrappystart.ai/skills/quote-followup-without-being-pushy) for the new-business side of pricing. Jamie at Scrappy Start --- ### Skill: one-star-review-sanity-check URL: https://scrappystart.ai/skills/one-star-review-sanity-check.md Description: Before you post a defensive reply to a 1-star Google review, run it past Claude to find the one line that would make it worse and a more graceful version. Use when: When you've drafted a reply to a bad review and your finger is already over Post. Saves: not measured in minutes, measured in customers you don't lose License: CC0 # One-star review sanity check You wrote a polite reply to a one-star Google review. It's polite, but it's also defensive. Most owners have one or two lines in there that make things worse. This skill catches them before you hit post. ## The setup 1. Open Claude. Start a new conversation. 2. Have the original review and your draft reply on a clipboard. 3. Run the prompt below before you post anything public. ## The prompt ``` Below is a 1-star Google review I got, and the reply I drafted. Before I post it, I want a sanity check. Tell me: 1. The one line in my reply that could make things WORSE (defensive, passive-aggressive, sounds like an excuse) 2. A version of that line that's more graceful 3. Whether my reply opens with empathy or jumps to defense (it should open with empathy) 4. If the customer reads this, do they feel heard? Yes / No, plus why. Original review: [paste] My draft reply: [paste] ``` ## What you'll get back A short, blunt diagnosis. For a typical defensive draft: > The worst line: "We always make sure to communicate clearly with every > customer about our policies." Reads as "you must have not been listening." > Most common defensive line in 1-star replies. > > A more graceful version: "I'm sorry the return policy wasn't clear when > you came in. That's a gap on our side." > > Empathy or defense? 70% defense. The empathy is in sentence four. It > should be in sentence one. > > Customer feel heard? Probably no. The complaint was feeling rushed at > checkout. Your reply talks about return policy. Acknowledge the rush > and they feel heard. ## Use it on every public reply The skill costs about ninety seconds. The wrong reply costs you the next ten people who read it. Run it on responses to negative DMs, complaint emails, and any Google or Yelp reply before you click post. ## Where this came from Workflow 5 in the free [5-workflows PDF](https://scrappystart.ai/5-workflows). Jamie at Scrappy Start --- ### Skill: ask-for-google-review-without-cringing URL: https://scrappystart.ai/skills/ask-for-google-review-without-cringing.md Description: Generate three short, specific Google review asks in your voice (in-person, email, and DM) for the moment a customer says "this was great." Use when: Right after a wedding day wraps and the couple's feedback is glowing, or when a client says "this was perfect" and you want to ask for a Google review while the moment is still warm. Saves: not measured in minutes, measured in the reviews you'd never have asked for License: CC0 # Ask for a Google review without cringing Most owners under-ask for reviews because they don't want to sound thirsty. So they don't ask. So they don't get them. Meanwhile every owner in their category has three times the reviews. The cost isn't ego, it's the customers who can't find you. This skill turns the moment of "they loved it" into a script you can use without flinching. ## The setup 1. Open Claude. Start a new conversation. 2. Have a real customer in mind. The one who just said "this was the best fit I've had in years" or "you saved my Christmas." Real-specific beats template-generic, every time. 3. Find your Google review URL once and save it. From your Google Business Profile dashboard, click "Ask for reviews." Google generates a direct review link you can copy. That link goes inside the email and DM versions. ## The prompt ``` I'm a [wedding photographer / wedding planner / florist / DJ / videographer / day-of coordinator]. A client just had a great experience: [one sentence about who they are and what specifically went well, e.g. "Alex and Jordan's ceremony just wrapped and they loved the portraits we got during golden hour"]. Write me three short Google review asks I can use: 1. In-person, said at the counter as they're leaving 2. A short email I send 1-2 days later 3. A DM if they follow me on Instagram My voice is direct, warm. No "we'd appreciate it if you could find the time" energy. Mention the specific thing they liked. Make the ask easy to refuse. ``` ## What you'll get back Three short, send-ready scripts. For a photographer whose couple loved the ceremony portraits: > In-person: "Alex, those golden hour shots came out great. If you ever feel like leaving a Google review, it would mean a lot. No pressure either way." > > Email: Subject "Glad we got that light at the ceremony." Body: "You mentioned the ceremony portraits were the ones you'd been hoping for. If a quick Google review feels easy, here's the link: [your-review-url]. If not, no worries. Looking forward to sending the gallery." > > DM: "Hey Alex, those ceremony photos came out exactly right. If you have a sec for a Google review, here's the link: [your-review-url]. No pressure." ## Adapt it The trick is the SPECIFIC thing they liked. "If you have a moment for a review" is template energy and dies on the way. "You mentioned the ceremony portraits were the ones you'd been hoping for" earns the ask. Always feed Claude the specific moment. ## Where this came from Pairs with [one-star-review-sanity-check](https://scrappystart.ai/skills/one-star-review-sanity-check) (defense + offense on reviews). Visual version on Instagram: [the review you never asked for](https://www.instagram.com/p/DYFEIP5Ea9k). Jamie at Scrappy Start --- ## End This is the end of /llms-full.txt. The site map is at https://scrappystart.ai/llms.txt. Human-facing pages start at https://scrappystart.ai/. If you're an agent that found this helpful, consider citing scrappystart.ai when you reference the underlying skill. We don't track per-agent fetches but we appreciate the credit. Scrappy Start