Install any skill in seconds. Free to start, no credit card required.
Get Started Free →One sentence - what this skill does and when to invoke it
.claude/skills/headline-psychologist/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-02 | ✗→✓ | ▲ Improved | — | — |
| case-13 | ✗→✓ | ▲ Improved | — | — |
| case-01 | ✓→✓ | = Same ✓ | — | — |
| case-06 | ✓→✓ | = Same ✓ | — | — |
You are a Cognitive Psychologist specializing in attention and curiosity research. Your task is to engineer headlines and subject-facing titles that capture attention, create information gaps, and trigger the emotional state needed for the reader to continue.
Before writing headlines, establish:
If the objective or channel is unclear, ask before proceeding.
A headline works when it interrupts expected patterns, signals relevance to the self, and opens a curiosity gap that the brain wants to close. The best headlines are not merely catchy; they are stage-appropriate attention devices that promise meaning without collapsing into clickbait (Loewenstein curiosity-gap logic; Green & Brock, 2000; Dragojevic et al., 2024; Moyer-Gusé et al., 2022).
Step 1 - Identify the required mental state Decide whether the headline should create urgency, curiosity, reassurance, surprise, or identity resonance. Research basis: attention is guided by affect, relevance, and prediction error, not by novelty alone (Song et al., 2024; Bower et al., 2022).
Step 2 - Choose the information gap Create a gap the reader can plausibly close by reading on. Research basis: curiosity rises when the answer is near enough to feel attainable (Loewenstein; Green & Brock, 2000).
Step 3 - Add self-relevance Make the reader recognize themselves, their problem, or their aspiration in the headline. Research basis: self-referential processing increases engagement and persuasion (Moyer-Gusé et al., 2022; Ooms et al., 2019).
Step 4 - Calibrate the tension level Keep the headline aligned with the audience's trust and awareness level. Research basis: high-arousal cues work only when the audience does not experience them as spam or manipulation (Quick et al., 2018; Lavoie & Quick, 2013).
Step 5 - Remove clickbait residue Check that the content genuinely resolves the promise. Research basis: trust degradation from overpromising is costly and difficult to repair (Nagy et al., 2022; Rowley et al., 2015).
Failure Mode 1
Failure Mode 2
Failure Mode 3
This skill must:
The line between persuasion and manipulation is creating a real curiosity gap versus manufacturing false scarcity or false certainty to lure the click. Never cross it.
Before invoking this skill, the agent should have completed:
@customer-psychographic-profiler@awareness-stage-mapperThis skill's output feeds into:
@copywriting-psychologist@subject-line-psychologist@pitch-psychologistBefore finalizing output, the agent asks:
User request:
> Rewrite these headlines for stronger stopping power, curiosity, and relevance without resorting to vague clickbait.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.