Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when the user wants a brief for one specific piece of content — angle, source insight, key message, 3 hook options, CTA, format, visual direction. Trigger on 'content brief', 'brief for this post', 'brief a writer', 'angle for this piece', 'brief for the LinkedIn post', or 'brief for a TikTok'.
.claude/skills/minhnv0807-36-content-brief-global/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 135% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 121% | 0% |
One brief per piece. This is the link between the calendar (01-content-calendar-global) and production (37-social-caption-global, 04-script-video-global, 05-ad-copy-global, 14-email-marketing-global). The brief is good enough when the writer finishes without asking the lead a single question.
Read .agents/product-marketing-context-global.md, the Brand Voice Document (35-brand-voice-global), and the calendar (01-content-calendar-global) first. If something is missing, ask up to 4 questions:
09-customer-insight-global: which internal monologue, pain, or objection is this piece aimed at.Take from the calendar (01-content-calendar-global): publish date, channel, format, funnel stage, pillar. Then set one — and only one — objective: reach / trust / conversion / engagement.
09-customer-insight-global): where is the reader in the journey — unaware of the problem, evaluating solutions, or close to buying.Recommend one with a short reason. The writer may choose differently if they can justify it.
[primary color], [secondary color] placeholders resolved from .agents/product-marketing-context-global.md), and any text overlay needed (a pulled quote or a number).42-image-brief-global, 43-carousel-brief-global, or 44-video-editor-brief-global.File name: content-brief-[piece-name]-[YYYYMMDD].md
markdown# Content Brief — [Internal title] ## Placement | Item | Detail | |------|--------| | Internal title | | | Channel | | | Format | video / carousel / static / text / email | | Funnel stage | TOFU / MOFU / BOFU | | Pillar | | | Target publish date | | | Approval deadline | | | Writer | | ## Angle and objective - **Source insight (from 09):** "[internal monologue / pain / objection — verbatim]" - **Angle:** [specific point of view — not a topic] - **Objective:** [exactly one — reach / trust / conversion / engagement] - **Reader's stage:** [unaware / evaluating / close to buying] ## Hooks — 3 options - **Hook A — Pain:** [...] - **Hook B — Curiosity:** [...] - **Hook C — Contrarian / surprise:** [...] - **Recommended:** [which one + short reason] ## Core content - **Key message (one sentence):** [...] - **Body (max 3 points):** 1. [Point 1] 2. [Point 2] 3. [Point 3] - **Evidence / social proof:** [specific number, testimonial, or case study + source] ## CTA - **Primary CTA:** [specific, performable] - **Micro-CTA (optional):** [comment keyword / save / share] - **Disclosure required:** [none / paid partnership / affiliate — exact wording] ## Brand voice - Tone: [reference Brand Voice Document, section II] - Banned words to avoid: [reference section III] ## Visual direction - **Concept:** [1-2 sentences] - **Dominant colors:** [from brand guideline] - **Text overlay:** [quote or number to place on the asset] - **Detailed visual brief:** hand off to `42-image-brief-global` / `43-carousel-brief-global` / `44-video-editor-brief-global` ## Distribution - **Publish time:** [per-channel peak] - **Hashtags:** [per platform norm] - **Cross-post:** [which channels, what changes] - **Paid amplification after publish:** Yes / No
01-content-calendar-global: supplies placement (date, channel, format, pillar) — runs before the brief.09-customer-insight-global: supplies the source insight, quoted verbatim into the brief.35-brand-voice-global: every brief cites tone and banned words from it.37-social-caption-global, 04-script-video-global, 05-ad-copy-global, 14-email-marketing-global: production skills that take this brief as input.42-image-brief-global, 43-carousel-brief-global, 44-video-editor-brief-global: detailed visual briefs, after copy approval.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 30,665 | 22,102 | -28% | 1 | 1 | 0% | 4,624 | 5,286 | +14% | 0 | 0 | — |
case-02 | fail→pass | 27,440 | 22,100 | -19% | 1 | 1 | 0% | 4,414 | 5,053 | +14% | 0 | 0 | — |
case-03 | fail→fail | 32,893 | 16,604 | -50% | 1 | 1 | 0% | 5,117 | 5,194 | +2% | 0 | 0 | — |
case-04 | pass→fail | 39,359 | 27,087 | -31% | 1 | 1 | 0% | 8,201 | 6,055 | -26% | 0 | 0 | — |
case-05 | fail→fail | 20,489 | 36,610 | +79% | 1 | 1 | 0% | 2,707 | 7,647 | +182% | 0 | 0 | — |
case-06 | pass→fail | 16,987 | 29,233 | +72% | 1 | 1 | 0% | 2,237 | 6,423 | +187% | 0 | 0 | — |
case-07 | fail→pass | 9,268 | 9,193 | -1% | 1 | 1 | 0% | 1,555 | 2,480 | +59% | 0 | 0 | — |
case-08 | fail→pass | 11,273 | 21,698 | +92% | 1 | 1 | 0% | 2,119 | 4,987 | +135% | 0 | 0 | — |
case-09 | pass→pass | 13,793 | 12,588 | -9% | 1 | 1 | 0% | 2,242 | 4,404 | +96% | 0 | 0 | — |
case-10 | fail→pass | 10,218 | 10,650 | +4% | 1 | 1 | 0% | 1,696 | 3,750 | +121% | 0 | 0 | — |
case-11 | fail→pass | 6,249 | 7,617 | +22% | 1 | 1 | 0% | 1,151 | 3,255 | +183% | 0 | 0 | — |
case-12 | pass→pass | 10,640 | 6,966 | -35% | 1 | 1 | 0% | 1,581 | 3,064 | +94% | 0 | 0 | — |
case-13 | fail→fail | 8,501 | 7,529 | -11% | 1 | 1 | 0% | 1,345 | 3,088 | +130% | 0 | 0 | — |
case-14 | pass→pass | 5,480 | 8,058 | +47% | 1 | 1 | 0% | 1,010 | 3,462 | +243% | 0 | 0 | — |
case-15 | pass→pass | 7,243 | 6,195 | -14% | 1 | 1 | 0% | 1,307 | 3,071 | +135% | 0 | 0 | — |
case-16 | pass→pass | 10,912 | 5,725 | -48% | 1 | 1 | 0% | 1,818 | 2,864 | +58% | 0 | 0 | — |
case-17 | fail→pass | 6,263 | 9,836 | +57% | 1 | 1 | 0% | 1,063 | 3,645 | +243% | 0 | 0 | — |
case-18 | fail→fail | 9,014 | 16,213 | +80% | 1 | 1 | 0% | 1,721 | 4,737 | +175% | 0 | 0 | — |
case-19 | fail→fail | 8,486 | 12,883 | +52% | 1 | 1 | 0% | 1,366 | 4,322 | +216% | 0 | 0 | — |
case-20 | fail→pass | 10,891 | 5,028 | -54% | 1 | 1 | 0% | 1,799 | 2,784 | +55% | 0 | 0 | — |
case-21 | fail→fail | 7,770 | 4,771 | -39% | 1 | 1 | 0% | 1,399 | 2,715 | +94% | 0 | 0 | — |
case-22 | pass→pass | 12,017 | 9,008 | -25% | 1 | 1 | 0% | 2,022 | 3,544 | +75% | 0 | 0 | — |
case-23 | fail→fail | 4,029 | 4,104 | +2% | 1 | 1 | 0% | 712 | 2,698 | +279% | 0 | 0 | — |
case-24 | fail→pass | 10,069 | 4,058 | -60% | 1 | 1 | 0% | 1,089 | 2,699 | +148% | 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. 24 cases were attempted. The headline lift of +29 percentage points is the difference between those two pass rates over the 24 comparable cases. 2 cases got worse with the skill loaded, and they are 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.