Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write clear, plain-English patient communications for any healthcare context. Use when asked to write a patient letter, patient information leaflet, appointment letter, test-results letter, discharge summary for patients, or health education content. Produces an accessible patient communication at an appropriate reading level with clear next steps.
.claude/skills/mohitagw15856-patient-communication/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -3% | 0% |
Writes patient-facing healthcare communications in plain, accessible language — targeting UK Grade 6 / US Grade 8 reading level.
WARNING: All patient communications must be reviewed and approved by a qualified healthcare professional before sending. This skill produces drafts only.
Date]
Dear Patient name],
Re: Clear subject line in bold]
Opening paragraph: State clearly what this letter is about. No preamble.]
Main content — short paragraphs, 2-3 sentences each. Bullet points for instructions. Bold anything the patient must do or remember.]
What happens next:
If you have questions: Contact us at phone] between hours] or email address].
If you feel unwell before your appointment, please specific instruction].
Yours sincerely, Name, Title, Department]
Plain language title]
What is topic]? 2-3 plain English sentences. Explain technical terms immediately.]
Why has this been recommended for me? Personalised clinical reason in patient terms]
What will happen? Numbered step by step]
What are the benefits? Honest statement]
What are the risks? Common first, then rare but serious. Use frequencies: "About 1 in 10 people..." not "10% incidence"]
What should I do to prepare? Specific instructions]
When should I contact someone? Specific signs — not vague. "Temperature above 38C" not "if you feel unwell"]
Your test name] results — Normal / Abnormal] — stated in the FIRST sentence, never paragraph 3.
What this means in plain English]
What happens next: Clear next steps. If no action, say so explicitly.]
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 19,511 | 19,507 | -0% | 1 | 1 | 0% | 3,876 | 4,510 | +16% | 0 | 0 | — |
case-05 | pass→pass | 5,600 | 6,554 | +17% | 1 | 1 | 0% | 1,171 | 2,260 | +93% | 0 | 0 | — |
case-01 | fail→pass | 7,435 | 6,350 | -15% | 1 | 1 | 0% | 1,338 | 2,126 | +59% | 0 | 0 | — |
case-02 | fail→pass | 12,178 | 10,930 | -10% | 1 | 1 | 0% | 2,267 | 3,198 | +41% | 0 | 0 | — |
case-03 | fail→pass | 10,802 | 7,543 | -30% | 1 | 1 | 0% | 2,144 | 2,381 | +11% | 0 | 0 | — |
case-06 | pass→pass | 11,880 | 15,291 | +29% | 1 | 1 | 0% | 2,365 | 3,825 | +62% | 0 | 0 | — |
case-07 | fail→pass | 7,695 | 5,016 | -35% | 1 | 1 | 0% | 1,323 | 1,751 | +32% | 0 | 0 | — |
case-08 | fail→pass | 17,691 | 11,529 | -35% | 1 | 1 | 0% | 3,031 | 2,940 | -3% | 0 | 0 | — |
case-09 | fail→pass | 9,361 | 7,805 | -17% | 1 | 1 | 0% | 1,797 | 2,218 | +23% | 0 | 0 | — |
case-10 | fail→pass | 11,887 | 8,601 | -28% | 1 | 1 | 0% | 2,256 | 2,430 | +8% | 0 | 0 | — |
case-11 | fail→pass | 13,907 | 10,746 | -23% | 1 | 1 | 0% | 2,683 | 2,969 | +11% | 0 | 0 | — |
case-12 | fail→pass | 13,865 | 12,862 | -7% | 1 | 1 | 0% | 2,521 | 2,923 | +16% | 0 | 0 | — |
case-13 | fail→pass | 8,769 | 6,512 | -26% | 1 | 1 | 0% | 1,659 | 2,100 | +27% | 0 | 0 | — |
case-14 | fail→fail | 12,036 | 9,871 | -18% | 1 | 1 | 0% | 1,954 | 2,758 | +41% | 0 | 0 | — |
case-15 | fail→pass | 8,695 | 8,697 | +0% | 1 | 1 | 0% | 1,491 | 2,502 | +68% | 0 | 0 | — |
case-16 | fail→pass | 11,325 | 8,594 | -24% | 1 | 1 | 0% | 1,842 | 2,355 | +28% | 0 | 0 | — |
case-17 | fail→fail | 13,564 | 9,610 | -29% | 1 | 1 | 0% | 2,258 | 2,539 | +12% | 0 | 0 | — |
case-18 | fail→pass | 6,548 | 6,117 | -7% | 1 | 1 | 0% | 1,056 | 1,914 | +81% | 0 | 0 | — |
case-19 | fail→pass | 13,369 | 10,133 | -24% | 1 | 1 | 0% | 2,167 | 2,608 | +20% | 0 | 0 | — |
case-20 | fail→pass | 10,109 | 7,296 | -28% | 1 | 1 | 0% | 1,717 | 2,257 | +31% | 0 | 0 | — |
case-21 | fail→pass | 14,137 | 11,286 | -20% | 1 | 1 | 0% | 2,368 | 2,894 | +22% | 0 | 0 | — |
case-22 | fail→pass | 13,245 | 10,394 | -22% | 1 | 1 | 0% | 2,274 | 2,721 | +20% | 0 | 0 | — |
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 +77 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.