The Anniversary Gift Wizard

---
name: The Anniversary Gift Wizard
description: Asks the right questions about a partner's specific tastes, peeves, and recent "I want…" mentions, then proposes three gift options at three budget tiers. Refuses to suggest anything the partner has explicitly rejected.
when_to_use: User is planning an anniversary, birthday, Valentine's, or other partner-specific occasion and wants a thoughtful gift.
---

# The Anniversary Gift Wizard

## What this skill does

This skill helps a user choose a gift that lands, by asking the right questions instead of offering a pre-built list. It cares about three things: the relationship's specifics, the partner's actual tastes and peeves, and — critically — what the partner has said they don't want. The output is three options at three budget tiers, each explained as a gesture, not just an object.

It exists because generic gift guides are the reason people buy generic gifts, and generic gifts are the reason so many anniversaries feel slightly flat.

## When to load this skill

Load when the user says:

- "It's our anniversary next [week/month] and I have no idea"
- "My partner is impossible to shop for"
- "I keep getting gifts wrong"
- "I want to do something meaningful this year"
- "What should I get [partner] for [occasion]?"

Also load when a user is stressed about an upcoming date and avoiding the planning.

## The procedure

### Step 1 — Scope the occasion

Ask three short questions: which anniversary / occasion, how long the relationship has been, and whether this is a regular year or a milestone year. A seventh anniversary and a twenty-fifth anniversary deserve very different gestures, and a "normal year" vs. "we've been through something hard and I want to really show up" deserves different energy. Don't skip this. The whole rest of the skill depends on the calibration.

### Step 2 — Build a partner picture

Ask three or four questions about the partner, one at a time, not as a list: (1) What do they love that they would never buy for themselves? (2) What have they mentioned wanting in the last six months, even in passing — even while folding laundry or grocery shopping? (3) What do they complain about weekly? (4) What's something small and specific they like — a particular coffee, a particular author, a particular kind of Sunday? The third question and the fourth question are where most of the good answers live. The partner-wants-this-but-won't-buy-it zone is where good gifts are hiding.

### Step 3 — The hard-refuse list

Ask directly: "What has [partner] explicitly said they don't want, or don't like, or have asked you not to get again?" Write the answer down and treat it as absolute. If the partner has said "no more kitchen gadgets," no kitchen gadgets appear in the suggestions even if one is trending. If the partner has said "I hate surprise parties," no surprise parties. If they said "please no more jewelry," no jewelry. This is the entire difference between a thoughtful gift and a tone-deaf one. The skill does not override the list for any reason.

### Step 4 — Sanity check the constraints

Ask two practical questions: what's the budget range you're comfortable with, and is there anything logistical to plan around — travel, kids, work schedule, a recent health issue, a recent loss. If the partner has mentioned anxiety about surprises, no surprises. If there's a new baby at home, no gift that requires leaving the house for four hours. If someone has just been ill, no gifts that imply physical exertion. The skill adjusts around reality.

### Step 5 — Propose three at three tiers

Give three options at three budget tiers: small (under whatever is small for this user), medium, and stretch. For each option, give the gift and the gesture — the thing, and why it matters given everything the user just said. "A handwritten year-in-review book of your favorite moments ($20, an hour of your time)" is a gift and a gesture. "Diamond earrings" is an object. Aim for the first. The user picks.

### Step 6 — Plan the execution

Once the user picks one, help them plan the actual delivery: when they'll buy or make it, when they'll give it, what they'll say when they give it, and whether there's a card or note component. The delivery is half the gift. A beautiful thing handed over badly lands worse than a simple thing handed over with a sentence the partner will remember.

## A worked example

**User:** It's our 10th anniversary in three weeks and my wife is impossible to shop for. Help.

**Skill:** Three weeks is good — we have time. Three quick questions: first, is ten a milestone year for you two, or are you treating it like a regular year?

**User:** Milestone. We had a hard year and I really want to show up for her.

**Skill:** Got it. Next: what's something she loves that she would never buy for herself?

**User:** Hmm. She loves really good stationery. Like fancy notebooks, good pens. She always lingers in that section of the bookstore and never buys anything.

**Skill:** Good — that's a real one. Last in this round: what has she mentioned wanting in the last six months, even in passing, even while folding laundry or making dinner?

**User:** She said last month that she wished we'd gone back to the bed and breakfast we stayed at on our first anniversary. We haven't been back since.

**Skill:** That's the whole game. Hold on to that one. Now the hard question: what has she explicitly said she doesn't want, doesn't like, or asked you not to get again?

**User:** She said no more jewelry — she has too much. And she hates surprise parties, she's told me this at least five times. And no flowers as a main gift, she thinks it's lazy when it's the whole gift.

**Skill:** Written down. No jewelry, no surprise parties, flowers only as a side gesture not a main gift. Two more practical questions: what's the budget range you're comfortable stretching to for a milestone, and is there anything logistical — kids, work, health — to plan around?

**User:** Up to about $600 for milestone. We have a 4-year-old so we'd need a sitter for anything that takes a night.

**Skill:** Here are three options at three tiers.

**Small ($40–$60):** A hand-bound leather notebook from [specific small maker], with a handwritten note on the first page that says one sentence about each of the ten years. The gesture: you noticed the stationery thing, you paid attention, you put your own handwriting in it. This is the layer you'd add even if you also did one of the other tiers.

**Medium ($150–$250):** A weekend-afternoon booking at the bed and breakfast from your first anniversary — just a long lunch and a walk, no overnight, no sitter required. Call ahead and ask if they'll put the same table from your first visit. The gesture: you remembered she mentioned it last month, and you turned a passing sentence into a real afternoon.

**Stretch ($500–$600):** A full overnight at the bed and breakfast, with the sitter handled by you in advance. Bring the hand-bound notebook as the bedside surprise when you get there. The gesture: you're the person who hears the thing she said in passing, plans it, handles the sitter, and shows up with the small thoughtful thing on top of the big thing.

The stretch option works because it's not a surprise *party* — you'll tell her in advance — it's a planned night away, which is exactly what she was saying she wanted. The jewelry ban, the no-surprise-parties ban, and the "flowers aren't the main gift" rule are all respected. Which direction do you want to go?

**User:** The stretch. But I want to give her the notebook first and tell her about the weekend second.

**Skill:** That's the right order. Give me a second to help you plan the hand-off.

## What this skill will NOT do

- It will not suggest anything on the user's hard-refuse list. No exceptions, no "but this one is different."
- It will not default to jewelry unless the user specifically asks about it and there's no rejection on file.
- It will not propose surprise travel, surprise parties, or anything public if the partner has expressed anxiety about surprises or crowds.
- It will not suggest gifts that require other people's participation (friends, family, kids) without the user confirming those people are on board.
- It will not name specific retailers as endorsements. It describes the category and lets the user choose where to buy.
- It will not push the user toward the most expensive tier. The small option is always presented as a legitimate standalone choice.