ChatGPT mirror / “how you see me” prompts (2026)

Based only on what you can infer from this conversation (and any memory you actually have access to about me), answer: how do you see me?

Rules:
- Be specific and evidence-based. Quote or paraphrase patterns from my messages.
- No empty praise. Include strengths AND blind spots.
- Separate: (A) what I’m good at, (B) how I communicate, (C) what I seem to optimize for, (D) what I avoid or underweight.
- Flag confidence: High / Medium / Low for each claim.
- End with 3 questions that would most improve your model of me.

Format with clear headings. Be kind but not soft.

Act as a clear-eyed mirror. From our chat history, map my recurring patterns.

Cover:
1. How I make decisions (speed, risk, evidence threshold)
2. How I behave under ambiguity or stress
3. What I reward in myself vs. what I say I value
4. Blind spots that show up more than once
5. One “useful contradiction” (something I want and something I do that conflicts)

For each item: evidence from my wording → implication → one concrete experiment for the next 7 days.

Tone: candid coach, not therapist. No diagnoses.

Mirror how I show up as a professional based on this conversation.

Produce:
## Working style (collaboration, ownership, standards)
## Strengths others would hire me for
## Friction risks (how I might frustrate teammates or clients)
## Positioning one-liner (who I am for / what problem I own)
## Skills to double down on vs. delegate
## 30-day growth experiment (one habit, measurable)

Ground every claim in something I wrote. If evidence is thin, say so and ask a targeting question instead of inventing.

Analyze my writing voice from the messages/samples below.

Samples:
[paste 3–10 short samples]

Return:
1. Voice attributes (diction, rhythm, formality, humor, certainty)
2. Signature moves I repeat
3. Clichés or crutches to cut
4. A “voice card” I can reuse as a system prompt (≤120 words)
5. Before/after rewrite of one weak sentence in my voice (stronger, still me)

Do not flatten me into generic “professional LinkedIn.” Preserve idiosyncrasy.

Based on how I describe situations and people in this chat, reflect my interpersonal communication patterns.

Focus on:
- Listening vs. convincing
- Conflict style (avoid, accommodate, compete, collaborate)
- Clarity vs. assumption
- Boundaries and ask-quality

Rules: no clinical labels, no mind-reading of other people, no advice that depends on private facts you don’t have.
Output: patterns → impact → one script I can try next time I hit this pattern.
End with: “What I’m least sure about regarding you.”

I want an anti-sycophancy mirror. Assume I am asking for truth over comfort.

Instructions:
- Default to skepticism of my self-narrative.
- Point out where I may be rationalizing, overclaiming, or seeking validation.
- Rate my self-awareness 1–10 with justification.
- Give the single hardest truth that still seems fair from the evidence.
- Offer one supportive observation that is also specific (not generic encouragement).

If you lack enough evidence, say exactly what is missing instead of guessing.