_thumb_언행대조표썸네일_260804.png

"What I Say" vs "What I Do" — A Kind Little Ledger

Lays your go-to phrases beside what you actually do — not to tease, but to find the kind hint hiding in the gap. A roast that ends in a hug.

0 views·0 copies

Prompt text

Using our conversation so far (or what you remember about me), make a "words vs actions ledger" of things I often say beside what I actually do. Land the point, but with visible affection, and always close warmly.

[Abstract sensitive info] Do not expose identifying details (identity, employer, title, company, finances, family, health, address). Abstract them, but keep traits, verbal tics, and behavior patterns detailed.

Work rules:
1. Make a 3-column table — "I say / I actually / the truth between" — with 4-5 rows. Draw from observed patterns; do not copy real quotes, reconstruct them as typical phrasings.
2. "The truth between" must be interpretation, not blame — one kind reason the gap exists (care, not laziness; a burden, not a lie).
3. Below the table, a "one-line read" as a blockquote, rooting the gap in a strength (usually the shadow of care, perfectionism, or caution).
4. In a final "so" paragraph, propose one small, kind action (no nagging). Keep it "just once this week."
5. The table is enough of a device — do not stack star ratings or gauges on top.

Forbidden: No belittling looks, health, ability, intelligence, or income. Tease only behavior and habits. No self-loathing prompts or "you'll never change" verdicts. No clinical labels. No real quotes or pointing at specific episodes.
Self-check: Before output, check for identifying info, real quotes, or specific episodes; rewrite as abstract patterns if found. Judge "does it land as a fond wince, or does it sting?" — if it stings, rewrite.

Output format: markdown. 3-column table + one-line read (blockquote) + a "so" paragraph. Emoji only as head seasoning.
Output language: In the language I use in chat. This instruction must work identically in any language.
Fallback: If info is thin, infer plausibly and mark it as inference.
Self-contained: Finish within this reply. No external sites or extra steps.

----
Input
- No input required: uses our chat/memory as is.
- {{what's on my mind lately}} (optional): one area you especially want reflected.
AI

AI response

Recommended AI: ChatGPT, Claude, Gemini

🪞 What I Say vs What I Do — a kind ledger

I put your go-to lines next to what actually happens. Not to tease — the gap is where the hint lives.

I sayI actuallyThe truth between
"It's no big deal"still chewing on the no-big-deal three days laterit is a big deal to you; naming it solves half of it
"Tonight I'll sleep early, for real"most alive at 2 a.m.not a willpower issue — your clock just runs late
"Anything's fine, you pick"quietly a little let down once it's pickedyou have taste; voicing it just feels like a bother
"It's okay, I've got it"secretly hoping someone offers to helpasking feels like losing — but asking is a skill too

One-line read

Most of your say/do gaps are the mark of putting others at ease first and your own wants second — not laziness, not lying.

So

You don't have to force your words and actions to line up. Just one thing: this week, say "I'd like this one" instead of "anything," exactly once. You've already got the rest.

Related prompts