Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Industry-adaptive B2B newsletter creation with stage, role, and geography-aware workflows
.claude/skills/brianrwagner-newsletter-creation-curation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-20 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 101% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 63% | 0% |
Use this skill to create B2B newsletters that match business context, not generic content templates.
Deep strategic guidance is in PLAYBOOK.md. Use this file as the executable operating manual.
Detect from context or ask: "Outline only, full edition, or full edition + calendar?"
| Mode | What you get | Best for | |------|-------------|----------| | quick | Edition outline + 3 lead story angles, no writing | Planning, editorial direction | | standard | Full newsletter edition, all sections written | Publishing this week | | deep | Full edition + 4-week content calendar + audience segmentation recommendations | Launching or systematizing a newsletter |
Default: standard — use quick if they're still deciding what to write. Use deep if they're building a newsletter from scratch or scaling.
Before generating any newsletter content, collect:
Structured intake — answer all 5 dimensions before proceeding:
Goal: [lead_gen | thought_leadership | personal_brand | category_ownership]
Industry: [sales_tech | hr_tech | fintech | ops_tech]
Stage: [series_a | series_b | series_c_plus]
Role: [founder | vp_director | pmm_content | enterprise_employee]
Geography: [us_first | india_first]Before drafting, reason through:
Output one-line strategy statement: > For [ICP], we publish [cadence] to achieve [goal] with [format].
Before writing full draft, produce an issue blueprint:
markdown## Issue Blueprint **Strategy:** For [ICP], [cadence] to achieve [goal]. **Industry tone:** [tactical/research-led/compliance-aware/practical] **Approval required:** [yes/no — who] **Sections:** 1. Subject lines (3 options) — [~words each] 2. Hook — [target ~75 words] 3. Core insight — [target ~200 words] 4. Actionable playbook — [target ~150 words, 3-5 steps] 5. CTA — [1 sentence, singular action]
Get confirmation or proceed to draft.
Generate complete publish-ready draft. Required structure — every issue:
Produce 3 distinct angles:
Run before delivering:
Clarity: Can a busy reader extract value in 60 seconds?Specificity: Does each section include concrete guidance or evidence?Relevance: Does tone match industry and role constraints?Compliance: For fintech/employee-led, is a legal/manager review step included?Consistency: Does voice align with prior issues (if any were provided)?CTA: Is there exactly ONE measurable CTA — not two, not zero?After completing the draft, evaluate:
Flag any issue: "The playbook steps are too generic for a Series B SaaS audience — they read as beginner content. Revised to assume existing process maturity."
After delivering the draft:
markdown## Newsletter Issue: [Name] — Issue #[X] — [Date] **Strategy:** For [ICP], [cadence] to achieve [goal]. --- ### Subject Line Options A) [Curiosity/open loop] B) [Specific + benefit] C) [Contrarian/bold] **Recommended:** [A/B/C] — [reason] --- ### Hook [2-3 sentences — problem + stakes] --- ### Core Insight [200-300 words — data, framework, or named pattern] --- ### Actionable Playbook 1. [Step] 2. [Step] 3. [Step] --- ### CTA [Single measurable action] --- ### Distribution Plan - **Send time:** [Day, Time, Timezone] - **LinkedIn amplification:** [Post angle / hook] - **Secondary channel:** [Platform + angle] ### KPI Targets - Open rate goal: [%] - CTR goal: [%] - Primary metric: [SQLs / subscribers / replies] ### Approval Required [Yes — [who] | No] ### Self-Critique Notes [Issues flagged + revisions made]
PLAYBOOK.md)SECTION ASECTION BSECTION CSECTION DCROSS-CUTTING: UNIVERSAL FRAMEWORKSSkill by Brian Wagner | AI Marketing Architect | brianrwagner.com
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-20 | fail→pass | 18,625 | 13,894 | -25% | 1 | 1 | 0% | 2,562 | 3,902 | +52% | 0 | 0 | — |
case-21 | pass→pass | 17,095 | 15,845 | -7% | 1 | 1 | 0% | 2,339 | 4,008 | +71% | 0 | 0 | — |
case-22 | pass→pass | 22,994 | 17,469 | -24% | 1 | 1 | 0% | 3,290 | 4,364 | +33% | 0 | 0 | — |
case-23 | pass→pass | 16,014 | 12,413 | -22% | 1 | 1 | 0% | 2,362 | 3,609 | +53% | 0 | 0 | — |
case-02 | fail→pass | 15,466 | 10,042 | -35% | 1 | 1 | 0% | 2,310 | 3,471 | +50% | 0 | 0 | — |
case-01 | fail→pass | 23,305 | 19,619 | -16% | 1 | 1 | 0% | 3,263 | 4,526 | +39% | 0 | 0 | — |
case-10 | pass→pass | 18,262 | 17,568 | -4% | 1 | 1 | 0% | 2,578 | 4,300 | +67% | 0 | 0 | — |
case-11 | fail→pass | 16,511 | 18,292 | +11% | 1 | 1 | 0% | 2,274 | 4,567 | +101% | 0 | 0 | — |
case-12 | fail→pass | 18,150 | 15,987 | -12% | 1 | 1 | 0% | 2,458 | 4,008 | +63% | 0 | 0 | — |
case-03 | fail→pass | 38,161 | 23,423 | -39% | 1 | 1 | 0% | 5,121 | 5,182 | +1% | 0 | 0 | — |
case-04 | pass→pass | 14,231 | 15,322 | +8% | 1 | 1 | 0% | 2,177 | 4,150 | +91% | 0 | 0 | — |
case-05 | pass→pass | 11,935 | 14,544 | +22% | 1 | 1 | 0% | 1,995 | 3,938 | +97% | 0 | 0 | — |
case-06 | pass→pass | 11,718 | 11,090 | -5% | 1 | 1 | 0% | 1,906 | 3,465 | +82% | 0 | 0 | — |
case-07 | fail→pass | 18,117 | 15,540 | -14% | 1 | 1 | 0% | 2,406 | 4,130 | +72% | 0 | 0 | — |
case-08 | pass→pass | 29,136 | 15,066 | -48% | 1 | 1 | 0% | 2,648 | 4,089 | +54% | 0 | 0 | — |
case-09 | pass→pass | 15,287 | 14,931 | -2% | 1 | 1 | 0% | 2,161 | 4,078 | +89% | 0 | 0 | — |
case-13 | fail→pass | 14,207 | 10,596 | -25% | 1 | 1 | 0% | 2,039 | 3,459 | +70% | 0 | 0 | — |
case-14 | fail→pass | 28,811 | 20,085 | -30% | 1 | 1 | 0% | 4,191 | 4,615 | +10% | 0 | 0 | — |
case-15 | fail→pass | 12,540 | 15,919 | +27% | 1 | 1 | 0% | 1,796 | 4,024 | +124% | 0 | 0 | — |
case-16 | pass→pass | 16,245 | 17,619 | +8% | 1 | 1 | 0% | 2,107 | 3,610 | +71% | 0 | 0 | — |
case-17 | pass→pass | 14,113 | 14,354 | +2% | 1 | 1 | 0% | 1,975 | 3,769 | +91% | 0 | 0 | — |
case-19 | fail→pass | 16,336 | 12,545 | -23% | 1 | 1 | 0% | 2,360 | 3,472 | +47% | 0 | 0 | — |
case-18 | fail→pass | 13,668 | 34,681 | +154% | 1 | 1 | 0% | 1,946 | 3,687 | +89% | 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. 23 cases were attempted. The headline lift of +52 percentage points is the difference between those two pass rates over the 23 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.