Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).
.claude/skills/itamarzand88-internal-comms/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 21% | 0% |
<!-- source: internal-comms — https://raw.githubusercontent.com/anthropics/skills/main/skills/internal-comms/SKILL.md -->
To write internal communications, use this skill for:
To write any internal communication:
examples/ directory:examples/3p-updates.md - For Progress/Plans/Problems team updatesexamples/company-newsletter.md - For company-wide newslettersexamples/faq-answers.md - For answering frequently asked questionsexamples/general-comms.md - For anything else that doesn't explicitly match one of the aboveIf the communication type doesn't match any existing guideline, ask for clarification or more context about the desired format.
3P updates, company newsletter, company comms, weekly update, faqs, common questions, updates, internal comms
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,618 | 7,317 | +11% | 1 | 1 | 0% | 1,206 | 1,428 | +18% | 0 | 0 | — |
case-06 | fail→pass | 14,808 | 8,761 | -41% | 1 | 1 | 0% | 2,873 | 1,991 | -31% | 0 | 0 | — |
case-02 | pass→pass | 8,336 | 7,757 | -7% | 1 | 1 | 0% | 1,274 | 1,590 | +25% | 0 | 0 | — |
case-03 | fail→pass | 8,520 | 8,620 | +1% | 1 | 1 | 0% | 1,419 | 1,713 | +21% | 0 | 0 | — |
case-04 | fail→fail | 7,835 | 8,077 | +3% | 1 | 1 | 0% | 1,494 | 1,669 | +12% | 0 | 0 | — |
case-05 | fail→fail | 9,805 | 10,509 | +7% | 1 | 1 | 0% | 2,006 | 2,231 | +11% | 0 | 0 | — |
case-07 | fail→fail | 8,535 | 10,209 | +20% | 1 | 1 | 0% | 1,616 | 2,132 | +32% | 0 | 0 | — |
case-08 | fail→fail | 8,514 | 8,856 | +4% | 1 | 1 | 0% | 1,519 | 1,957 | +29% | 0 | 0 | — |
case-09 | pass→pass | 9,142 | 7,022 | -23% | 1 | 1 | 0% | 1,887 | 1,798 | -5% | 0 | 0 | — |
case-10 | fail→fail | 5,253 | 7,428 | +41% | 1 | 1 | 0% | 922 | 1,841 | +100% | 0 | 0 | — |
case-11 | fail→fail | 9,913 | 7,623 | -23% | 1 | 1 | 0% | 1,980 | 1,706 | -14% | 0 | 0 | — |
case-12 | pass→pass | 8,192 | 6,632 | -19% | 1 | 1 | 0% | 1,711 | 1,539 | -10% | 0 | 0 | — |
case-13 | fail→fail | 10,384 | 6,842 | -34% | 1 | 1 | 0% | 2,276 | 1,603 | -30% | 0 | 0 | — |
case-14 | fail→fail | 5,718 | 6,276 | +10% | 1 | 1 | 0% | 1,110 | 1,457 | +31% | 0 | 0 | — |
case-15 | fail→pass | 8,678 | 6,221 | -28% | 1 | 1 | 0% | 1,593 | 1,492 | -6% | 0 | 0 | — |
case-16 | pass→pass | 7,772 | 5,611 | -28% | 1 | 1 | 0% | 1,389 | 1,293 | -7% | 0 | 0 | — |
case-17 | fail→pass | 10,440 | 12,085 | +16% | 1 | 1 | 0% | 1,772 | 2,373 | +34% | 0 | 0 | — |
case-18 | fail→pass | 7,881 | 7,062 | -10% | 1 | 1 | 0% | 1,338 | 1,622 | +21% | 0 | 0 | — |
case-19 | pass→pass | 9,655 | 10,528 | +9% | 1 | 1 | 0% | 2,113 | 2,240 | +6% | 0 | 0 | — |
case-20 | pass→pass | 9,561 | 8,131 | -15% | 1 | 1 | 0% | 1,878 | 1,955 | +4% | 0 | 0 | — |
case-21 | pass→pass | 14,029 | 8,499 | -39% | 1 | 1 | 0% | 2,953 | 2,176 | -26% | 0 | 0 | — |
case-22 | pass→pass | 9,600 | 8,987 | -6% | 1 | 1 | 0% | 1,862 | 1,863 | +0% | 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 +23 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.