ads
無料Generates paid ad copy for Google, Meta, and LinkedIn. Produces platform-ready variants using a creative matrix approach with headlines, descriptions, and image briefs.
日本語の概要は準備中です。原文の説明を表示しています。
Interactive workshop that builds a monthly rolling editorial calendar using MKT1 methodology. Walks through Perceptions, content pillars, themes, and capacity to produce a 4-week detailed plan plus 3-month roadmap. Designed to be re-run monthly.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
This skill runs an interactive workshop to build a monthly editorial calendar grounded in Emily Kramer's MKT1 methodology. It walks through five discovery sections, captures answers, and generates a rolling content plan.
The output is designed to be re-run monthly: a detailed 4-week editorial calendar with specific content pieces, channels, and owners, plus a high-level 3-month theme roadmap that shifts forward each time.
Discovery Mode (Steps 1 through 5): Walk through five sections of questions. Capture answers as they come. After each section, summarize what you heard and confirm before moving on. Do not invent answers. Do not fill gaps with assumptions. If a question goes unanswered, flag it as [OPEN].
Generate Mode (Step 6): When all discovery sections are complete (or the user says "generate it" or "build the calendar"), take everything captured and produce the editorial calendar output.
Refresh Mode: If the user already has a prior calendar and says "refresh" or "update the calendar," ask only what has changed since last month (new goals, new product launches, performance data) and regenerate. Skip sections where nothing changed.
Why we start here: The editorial calendar exists to serve business goals, not to fill a publishing schedule. Emily Kramer's first principle: avoid random acts of marketing.
Questions to walk through:
What to capture: The strategic frame that every content decision filters through. If content does not serve one of these goals, it should not be on the calendar. Note which revenue lever(s) are the priority so the calendar can weight content accordingly.
Why we go here next: Perceptions are what you want your audience to repeat about you. They are the filter for every piece of content. If a piece does not reinforce at least one Perception, reconsider creating it.
Questions to walk through:
What to capture: 3 to 5 Perceptions, ranked by priority. The gap between current state and desired state for each. The evidence inventory (what proof exists, what needs to be created). These become the backbone of the content pillar structure.
Why we go here next: Pillars translate Perceptions into content categories. Each pillar maps to one or more Perceptions and addresses the needs of a specific audience segment.
Questions to walk through:
What to capture: Audience segments with channel and format preferences. 3 to 5 content pillars mapped to Perceptions. Current channel mix and cadence. This determines what types of content populate the calendar and where they get distributed.
Why we go here: A calendar that exceeds capacity is fiction. Emily Kramer's rule: default to less, not more. Do not sacrifice quality bar and distribution for volume.
Questions to walk through:
What to capture: Team capacity (pieces/week by format). Content velocity constraints. What topics and formats have proven traction. Repurposable asset inventory. This sets the realistic volume for the calendar.
Why we go here last: Product and market events create the "tent poles" around which the rest of the calendar is organized.
Questions to walk through:
claude.md: major = human element required, medium = automate as much as reasonable, minor = fully automated.What to capture: A timeline of external and internal events that anchor the calendar. Launch tier for each product release. Conference and market moment dates. These become the fixed points around which flexible content is scheduled.
When all five sections are complete (or the user triggers generation), produce the editorial calendar.
Theme assignment:
Content selection:
Repurposing track:
Distribution plan:
Reactive buffer:
Generate three artifacts:
## 3-Month Theme Roadmap
| Month | Theme | Revenue Lever | Perception Focus | Tent Pole Events | Anchor Content |
|-------|-------|---------------|------------------|------------------|----------------|
| [Month 1] | [Theme name] | [Which lever] | [P1, P3] | [Product launch, conference] | [1-2 big pieces] |
| [Month 2] | [Theme name] | [Which lever] | [P2, P4] | [Events] | [1-2 big pieces] |
| [Month 3] | [Theme name] | [Which lever] | [P1, P5] | [Events] | [1-2 big pieces] |
### Month 1: [Theme Name]
**Why this theme:** [1-2 sentences connecting to business goals and market context]
### Month 2: [Theme Name]
**Why this theme:** [1-2 sentences]
### Month 3: [Theme Name]
**Why this theme:** [1-2 sentences]
## 4-Week Editorial Calendar: [Month Name Year]
### Week 1: [Date Range]
| Day | Content | Type | Pillar | Perception | Channel | Distribution | Owner | Status |
|-----|---------|------|--------|------------|---------|-------------|-------|--------|
| Mon | [Title/Topic] | Blog | [Pillar] | P1 | Website | LinkedIn, Newsletter, X | [TBD] | Draft |
| Wed | [Title/Topic] | Social | [Pillar] | P3 | LinkedIn | X cross-post | [TBD] | Draft |
| Thu | [Title/Topic] | Newsletter | [Pillar] | P1, P2 | Email | LinkedIn teaser | [TBD] | Draft |
### Week 2: [Date Range]
[Same table structure]
### Week 3: [Date Range]
[Same table structure]
### Week 4: [Date Range]
[Same table structure]
### Reactive Buffer
**Reserved capacity:** [X pieces/week] for:
- Trending topics in [audience channels]
- Competitive responses
- Community questions from [forums/channels]
- Breaking news in [industry area]
For each content piece in the calendar, include a one-line brief:
## Content Briefs
### [Content Title 1]
**Type:** [Blog/Social/Newsletter/Changelog/Tutorial]
**Pillar:** [Which content pillar]
**Perception:** [Which perception(s) this reinforces]
**Angle:** [The specific argument or hook. Not just the topic, the point of view.]
**Target audience:** [Which segment]
**Funnel stage:** [Awareness/Consideration/Decision/Expansion]
**Distribution:** [Primary channel] → [Amplification channels]
**Pre-post opportunity:** [Yes/No. If yes, suggest the teaser.]
**Repurposing:** [What derivatives can be created from this piece]
## Repurposing Matrix
| Source (Anchor Content) | Derivative 1 | Derivative 2 | Derivative 3 | Timing |
|------------------------|-------------|-------------|-------------|--------|
| [Blog: Topic] | LinkedIn post | Newsletter section | X thread | Same week |
| [Tutorial: Topic] | Short video clip | Code snippet social | Docs update | +1 week |
| [Changelog: Feature] | Social announcement | Blog deep dive | Email to customers | Same day |
Before delivering, verify the calendar passes:
When the user is pre-revenue or very early stage:
When the user already has a messaging_positioning.md:
When the user says "refresh" or "update":
When capacity is very low (1 piece/week or less):
When the user cannot articulate Perceptions:
When product roadmap is unclear:
After the calendar is generated, ask:
"Would you like to start generating any of these content pieces now? I can hand off to:"
/blogfor blog posts and tutorials/social-postsfor social media content/launch-roundupfor product launch packages
If the user wants to save the calendar for reference:
"I'll save this to
output/editorial-calendar/[month-year]-editorial-calendar.md. Re-run/editorial-calendarnext month with 'refresh' to roll the plan forward."
This skill is built on Emily Kramer's MKT1 frameworks:
Source: MKT1 Newsletter and MKT1 Field Guide to B2B Startup Marketing
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Generates paid ad copy for Google, Meta, and LinkedIn. Produces platform-ready variants using a creative matrix approach with headlines, descriptions, and image briefs.
日本語の概要は準備中です。原文の説明を表示しています。
Use when you want paid ad copy for Google, Meta, or LinkedIn, including headlines, descriptions, and image briefs.
日本語の概要は準備中です。原文の説明を表示しています。
Use when you want campaign performance reviewed against benchmarks, with a health score and prioritized optimization recommendations.
日本語の概要は準備中です。原文の説明を表示しています。
Use when you want a strict review of a marketing asset against style rules, factual accuracy, and messaging alignment.
日本語の概要は準備中です。原文の説明を表示しています。
Skill Optimization: Auto Research. Autonomously optimize any Claude Code skill by running it repeatedly, scoring outputs against binary evals, mutating the prompt, and keeping improvements. Based on Karpathy's autoresearch methodology. Use when: optimize this skill, improve this skill, run autoresearch on, make this skill better, self-improve skill, benchmark skill, eval my skill, run evals on. Outputs: an improved SKILL.md, a results log, and a changelog of every mutation tried.
日本語の概要は準備中です。原文の説明を表示しています。
Use when you want to benchmark and improve an existing repo skill through repeated evals, prompt mutations, and score tracking.
日本語の概要は準備中です。原文の説明を表示しています。