Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Develop a content strategy covering editorial positioning, content pillars, formats, calendar, governance, and topical authority planning. Use this skill whenever the user wants to plan a content program, define content pillars, build an editorial calendar, structure topic clusters, set up content governance, or align content production with broader brand and SEO goals. Triggers on content strategy, content plan, editorial strategy, content pillars, content calendar, editorial calendar, topical
.claude/skills/rampstackco-content-strategy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 30% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 53% | 0% |
Plan what content to produce, why, when, and how. Stack-agnostic. Works for blogs, knowledge bases, marketing sites, newsletters, and product content surfaces.
This skill is the strategic layer. Tactical execution sits in content-and-copy, landing-page-copy, and email-sequences. SEO-driven content planning sits in seo-keyword. This skill stitches those together into a coherent program.
content-and-copy)landing-page-copy)email-sequences)seo-keyword, then return here for strategic planning)seo-content-audit)brand-discovery or creative-brief)seo-keyword) if SEO-drivenIf brand positioning is unclear, run brand-discovery first. If audience is undefined, do that work before strategy.
A content strategy has five layers. Skip layers and the program drifts.
What this content program stands for. The "why we write" statement.
Components:
Example structure:
> We help audience] achieve outcome] by publishing content type] that unique angle], unlike common alternatives] that common shortcoming].
A strong editorial positioning forces choices. If you cannot complete the sentence, the positioning is too vague.
The 3 to 5 themes the program owns. Every piece of content belongs to a pillar.
Per pillar, define:
Pillar selection criteria:
3 to 5 pillars is the sweet spot. Fewer than 3 risks brittleness. More than 5 dilutes focus.
The shapes the content takes.
Format dimensions:
Format selection criteria:
A content program typically has 3 to 5 formats it returns to consistently. More than that fragments production and brand recognition.
When content publishes.
Cadence options:
Calendar structure:
Common failure: publishing whatever happens to get written. Without a calendar, the program drifts toward the topics easiest to write, not the ones the audience needs most.
How content gets made, reviewed, published, measured, and retired.
Production workflow:
Idea → Brief → Outline → Draft → Edit → Review → Publish → Measure → Update or RetirePer stage, define:
Roles:
Lifecycle decisions:
seo-content-audit).references/content-strategy-template.md.Default output is a strategy document at content-strategy.md plus an editorial calendar in whatever tool the team uses.
Strategy document structure:
Editorial calendar (separate, ongoing):
This skill's output depends on data, measurements, or tool results it cannot generate on its own. When a required input, tool, or data source is unavailable or unverifiable, the sanctioned output is the deliverable with the gap stated: what was needed, what was actually obtained or verified, and which parts of the output are affected. Fabricating, estimating, or interpolating a required number to complete the deliverable is never sanctioned. A stated gap is a complete answer.
references/content-strategy-template.md - Strategy document template.references/editorial-calendar-template.md - Spreadsheet column definitions and calendar structure.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | pass→pass | 13,483 | 9,908 | -27% | 1 | 1 | 0% | 2,278 | 3,974 | +74% | 0 | 0 | — |
case-01 | fail→fail | 40,817 | 31,774 | -22% | 1 | 1 | 0% | 6,242 | 7,885 | +26% | 0 | 0 | — |
case-02 | fail→fail | 36,646 | 30,302 | -17% | 1 | 1 | 0% | 5,976 | 7,348 | +23% | 0 | 0 | — |
case-03 | fail→pass | 38,042 | 28,667 | -25% | 1 | 1 | 0% | 6,238 | 7,163 | +15% | 0 | 0 | — |
case-04 | fail→fail | 35,328 | 25,705 | -27% | 1 | 1 | 0% | 6,197 | 6,675 | +8% | 0 | 0 | — |
case-05 | fail→fail | 15,824 | 11,856 | -25% | 1 | 1 | 0% | 2,390 | 3,937 | +65% | 0 | 0 | — |
case-06 | fail→pass | 20,717 | 12,608 | -39% | 1 | 1 | 0% | 3,168 | 4,175 | +32% | 0 | 0 | — |
case-07 | fail→fail | 17,613 | 15,341 | -13% | 1 | 1 | 0% | 2,639 | 4,669 | +77% | 0 | 0 | — |
case-08 | fail→pass | 21,399 | 13,909 | -35% | 1 | 1 | 0% | 3,271 | 4,240 | +30% | 0 | 0 | — |
case-09 | fail→pass | 20,202 | 16,936 | -16% | 1 | 1 | 0% | 3,285 | 4,943 | +50% | 0 | 0 | — |
case-10 | fail→fail | 5,830 | 6,453 | +11% | 1 | 1 | 0% | 889 | 3,226 | +263% | 0 | 0 | — |
case-11 | fail→pass | 20,767 | 18,100 | -13% | 1 | 1 | 0% | 3,310 | 5,075 | +53% | 0 | 0 | — |
case-12 | fail→fail | 17,814 | 19,625 | +10% | 1 | 1 | 0% | 2,672 | 5,196 | +94% | 0 | 0 | — |
case-13 | pass→pass | 28,951 | 28,116 | -3% | 1 | 1 | 0% | 4,687 | 6,664 | +42% | 0 | 0 | — |
case-14 | pass→pass | 15,351 | 17,886 | +17% | 1 | 1 | 0% | 2,420 | 5,100 | +111% | 0 | 0 | — |
case-15 | pass→pass | 13,266 | 9,837 | -26% | 1 | 1 | 0% | 2,026 | 3,639 | +80% | 0 | 0 | — |
case-17 | fail→fail | 16,185 | 14,380 | -11% | 1 | 1 | 0% | 2,902 | 4,715 | +62% | 0 | 0 | — |
case-18 | fail→pass | 15,402 | 14,408 | -6% | 1 | 1 | 0% | 2,341 | 4,436 | +89% | 0 | 0 | — |
case-19 | pass→pass | 16,091 | 14,659 | -9% | 1 | 1 | 0% | 2,324 | 4,347 | +87% | 0 | 0 | — |
case-20 | pass→pass | 16,914 | 14,399 | -15% | 1 | 1 | 0% | 2,708 | 4,312 | +59% | 0 | 0 | — |
case-21 | pass→pass | 17,378 | 11,507 | -34% | 1 | 1 | 0% | 2,388 | 3,759 | +57% | 0 | 0 | — |
case-22 | fail→pass | 18,702 | 12,551 | -33% | 1 | 1 | 0% | 2,840 | 4,162 | +47% | 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 +32 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.