Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Plan multi-platform social campaigns: briefs to posting.
.claude/skills/nousresearch-social-media-content-calendar/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 166% | 0% |
| case-06 | ✓→✓ | = Same ✓ | -27% | 0% |
| case-08 | ✓→✓ | = Same ✓ | 12% | 0% |
| case-09 | ✓→✓ | = Same ✓ | 147% | 0% |
Plan a concrete calendar across selected social platforms. This skill owns campaign structure, post briefs, channel adaptation, approvals, and publishing verification; platform skills such as xurl own API commands. For platforms without a connector, the verified handoff ends at approved drafts for the user's scheduler — say so rather than claiming publication.
Don't use for: single one-off posts (use the platform skill directly).
Record objective, audience, offer/message, platforms, date range, cadence, voice, mandatory/prohibited claims, links, tracking convention, localization, and approval/publishing authority. Done when each proposed post has a clear business purpose.
Collect verified product facts, launches, articles, media, testimonials with permission, brand assets, and key dates using read_file and web_extract. Mark claim owners and expiration. Done when unsupported claims and missing assets are visible.
Create a balanced mix such as education, proof, product, community, event, behind-the-scenes, and conversation. Account for platform cadence and campaign milestones. Done when dates, platforms, themes, and objectives form a coherent calendar rather than duplicate cross-posts.
For each post specify hook, core message, format, copy length, CTA, link, asset dimensions/content, accessibility text, tags/mentions, and success metric. Adapt rather than copy-paste between platforms. Done when a creator can produce the asset without hidden context.
Load humanizer for voice; generate visuals with the image_generate tool where assets are needed. Preserve factual claims and shared campaign identity while respecting platform norms. Done when every calendar slot has draft copy and asset status.
Check factual accuracy, tone, repetition, rights/permissions, accessibility, disclosures, link destination, date relevance, and crisis sensitivity. Mark draft, needs review, or approved; do not publish from draft. Done when every post has a disposition and owner.
Present the approval batch. Publish/schedule only approved posts using available platform skills (xurl for X); for platforms without a connector, deliver the approved package (copy, assets, timing) for the user's scheduling tool and mark those slots handed-off, not published. Read back scheduled time, account, content preview, and provider post/job ID for anything actually published. Done when the calendar reflects verified publishing or handoff status per slot.
draft or needs review state.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 51,283 | 45,647 | -11% | 1 | 1 | 0% | 8,204 | 6,575 | -20% | 0 | 0 | — |
case-02 | fail→fail | 55,410 | 41,054 | -26% | 1 | 1 | 0% | 8,245 | 6,473 | -21% | 0 | 0 | — |
case-03 | fail→fail | 51,374 | 30,708 | -40% | 1 | 1 | 0% | 8,248 | 5,619 | -32% | 0 | 0 | — |
case-04 | pass→fail | 4,372 | 7,997 | +83% | 1 | 1 | 0% | 712 | 1,894 | +166% | 0 | 0 | — |
case-05 | fail→fail | 7,441 | 7,245 | -3% | 1 | 1 | 0% | 1,171 | 1,874 | +60% | 0 | 0 | — |
case-06 | pass→pass | 47,940 | 31,591 | -34% | 1 | 1 | 0% | 7,400 | 5,389 | -27% | 0 | 0 | — |
case-07 | fail→fail | 8,091 | 10,576 | +31% | 1 | 1 | 0% | 1,192 | 2,297 | +93% | 0 | 0 | — |
case-08 | pass→pass | 19,434 | 14,632 | -25% | 1 | 1 | 0% | 2,725 | 3,044 | +12% | 0 | 0 | — |
case-09 | pass→pass | 13,963 | 29,229 | +109% | 1 | 1 | 0% | 2,340 | 5,773 | +147% | 0 | 0 | — |
case-10 | pass→pass | 6,235 | 6,237 | +0% | 1 | 1 | 0% | 838 | 1,623 | +94% | 0 | 0 | — |
case-11 | fail→fail | 19,278 | 34,074 | +77% | 1 | 1 | 0% | 3,084 | 5,998 | +94% | 0 | 0 | — |
case-12 | fail→pass | 13,143 | 6,516 | -50% | 1 | 1 | 0% | 1,853 | 1,682 | -9% | 0 | 0 | — |
case-13 | fail→fail | 19,190 | 30,566 | +59% | 1 | 1 | 0% | 3,534 | 5,597 | +58% | 0 | 0 | — |
case-14 | fail→fail | 3,730 | 4,187 | +12% | 1 | 1 | 0% | 630 | 1,358 | +116% | 0 | 0 | — |
case-15 | fail→fail | 12,566 | 17,881 | +42% | 1 | 1 | 0% | 1,916 | 3,698 | +93% | 0 | 0 | — |
case-16 | pass→pass | 7,905 | 10,281 | +30% | 1 | 1 | 0% | 1,196 | 2,357 | +97% | 0 | 0 | — |
case-17 | pass→pass | 7,304 | 8,833 | +21% | 1 | 1 | 0% | 1,173 | 2,262 | +93% | 0 | 0 | — |
case-18 | pass→pass | 6,568 | 16,708 | +154% | 1 | 1 | 0% | 894 | 2,277 | +155% | 0 | 0 | — |
case-19 | fail→fail | 23,497 | 10,592 | -55% | 1 | 1 | 0% | 3,530 | 2,366 | -33% | 0 | 0 | — |
case-20 | pass→pass | 9,896 | 13,523 | +37% | 1 | 1 | 0% | 1,446 | 2,702 | +87% | 0 | 0 | — |
case-21 | pass→pass | 7,937 | 5,258 | -34% | 1 | 1 | 0% | 1,154 | 1,589 | +38% | 0 | 0 | — |
case-22 | fail→fail | 14,211 | 19,647 | +38% | 1 | 1 | 0% | 2,393 | 3,797 | +59% | 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 0 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.