code-ai-installer 3.0.0 → 4.0.1-a

package/domains/content/locales/en/.agents/skills/data-storytelling/SKILL.md

@@ -1,294 +1,309 @@
----
-name: data-storytelling
-description: Turning data into narratives — data story types, persuasion techniques, data-driven visualization, platform adaptation
----
-# Data Storytelling — Data-Driven Narratives
-
-## When to use
-- After **gathering facts and statistics** (`$topic-research`, `$source-verification`) — to structure them into a story.
-- When creating **analytical content** — articles, carousels, infographics, reports.
-- When **justifying a position** — case studies, comparisons, predictions.
-- To **reinforce any content** with data — even 1-2 numbers make a post more persuasive.
-
-## Input
-
-| Field | Required | Description |
-|------|:-----------:|----------|
-| Data / facts | ✅ | Numbers, statistics, research findings — the building blocks of the story |
-| Data sources | ✅ | Where the data comes from: research, analytics, API, report, `$source-verification` |
-| Topic / context | ✅ | What we are talking about — mapping to the content topic |
-| Target persona | ✅ | From `$audience-analysis` — dictates depth and language |
-| Platform | ✅ | Determines volume, visualization format, and depth |
-| Content goal | ⬚ | To inform / persuade / engage / convert |
-| Data story type | ⬚ | If omitted — the agent determines it based on the data's character |
-
-> If data is absent — **ask the user** or redirect to `$topic-research` / `$source-verification`. Do not fabricate numbers.
-
-## Data Story Types
-
-| Type | When to use | Question it answers | Example |
-|-----|-------------------|----------------------------|--------|
-| **Trend** | Data shows change over time | "What's happening and where is it heading?" | "The market for AI agents grew by 340% in a year" |
-| **Comparison** | Comparing 2+ subjects or periods | "Who is better / what has changed?" | "Email vs Telegram: Marketing channel ROI" |
-| **Anomaly** | An unexpected fact, an outlier, a contradiction | "Why is this strange?" | "87% of companies use AI, but only 12% see ROI" |
-| **Correlation** | Connection between two factors | "What is linked to what?" | "Companies with a content strategy generate 6× more leads" |
-| **Forecast** | Data allows for extrapolation | "What happens next?" | "By 2027, 50% of content will be AI-generated" |
-| **Scale** | Showcasing the magnitude of a phenomenon | "Just how big/small is this?" | "7.5M blog posts are published daily — that's 87 posts a second" |
-
-> The type dictates the narrative structure and choice of visualization.
-
-## Narrative structures by type
-
-### Trend
-```
-Context (what was) → Inflection point → Trend (data) → What it means → Forecast/CTA
-```
-
-### Comparison
-```
-Broad context → Subject A (data) → Subject B (data) → Key differentiator → Conclusion
-```
-
-### Anomaly
-```
-Expectation (what everyone thinks) → Reality (data) → Why it's the case → Insight → Conclusion
-```
-
-### Correlation
-```
-Observation → Connection (data A + data B) → Cause/mechanism → What to do about it
-```
-
-### Forecast
-```
-Current state (data) → Growth/decline factors → Forecast → What to do right now
-```
-
-### Scale
-```
-Abstract number → Scaling (translating to relatable terms) → Why it matters → Conclusion
-```
-
-## Persuasion techniques
-
-| Technique | Description | Example |
-|---------|----------|--------|
-| **Scaling** | Translating an abstract number into something relatable | "$3.5 trillion — that's more than the UK's GDP" |
-| **Simile / Comparison** | Juxtaposing with something familiar | "It's as if every third resident of London..." |
-| **Antithesis** | Contrasting expectation and reality | "Everyone thinks X. In reality — Y" |
-| **Metaphor** | Converting data into imagery | "Email marketing is like fishing: 95% of the catch depends on the bait (subject line)" |
-| **Personalization** | Bringing it down to a single person's level | "The average marketer spends 6 hours a week planning content" |
-| **Before/After** | Showing the state before and after | "Before automation: 3 days. After: 15 minutes" |
-| **Rule of Three** | Three facts are remembered best | "Faster. Cheaper. More efficient." |
-| **Q&A** | Asking a question, answering with data | "How much does a lead cost? $5 on Instagram vs $50 on LinkedIn" |
-
-> Use 1-3 techniques per data story. Any more results in overload.
-
-## Visualization by data type
-
-| Data / Story type | Recommended visualization | When NOT to use |
-|--------------------|--------------------------|-----------------------|
-| **Trend** (time) | Line chart, Area chart | Pie chart (lacks a time axis) |
-| **Comparison** (categories) | Bar chart (horizontal), Grouped bar | Line chart (implies continuity) |
-| **Proportions** (parts of whole)| Pie chart (≤ 5 segments), Stacked bar | Pie chart with 10+ segments |
-| **Correlation** (two vars) | Scatter plot, Bubble chart | Bar chart (doesn't show connection) |
-| **Scale** (one big fact) | Big number + context, Pictogram | Complex graph (kills the impact) |
-| **Ranking** (order) | Horizontal bar (sorted) | Pie chart (no inherent order) |
-| **Process** (stages) | Funnel chart, Sankey diagram | Pie chart |
-| **For social media** | Carousel w/ 1 graph per slide, infographics | Complex dashboards |
-
-> If data is sparse (1-3 numbers) — visualization is unnecessary. A prominent number within the text suffices.
-
-## Data Validation (mandatory)
-
-Prior to building a narrative — verify the data:
-
-| Criterion | Question | Risk if unverified |
-|----------|--------|----------------------|
-| **Recency** | When was the data gathered? | Outdated statistics = false conclusions |
-| **Source** | Who conducted the research? Is there a methodology? | Unreliable source = zero trust |
-| **Sample size** | How many respondents / data points? | Small sample = mere coincidence |
-| **Context** | Is the data cherry-picked? | Cherry-picking = manipulation |
-| **Causation vs Correlation**| Is the link causal or coincidental? | False causality = erroneous takeaway |
-| **Completeness** | Are both "pro" AND "con" data shown? | One-sidedness = bias |
-
-> If data fails validation — discard it. Fewer, yet reliable numbers are vastly superior.
-
-## Platform adaptation
-
-| Platform | Depth | Number of facts | Visualization | Format |
-|-----------|---------|-------------------|-------------|--------|
-| **LinkedIn** | High — comprehensive analysis | 5-10 facts | Infographics, carousel, graph screenshot | Long post / article |
-| **Instagram** | Medium — 1 insight + 2-3 facts | 2-4 facts | Carousel (1 fact = 1 slide), infographics | Carousel / reel |
-| **Telegram** | Medium-to-High | 3-7 facts | Embedded tables, link to full report | Well-formatted post |
-| **Twitter/X** | Low — 1 hook-fact | 1-2 facts | Graph screenshot / big number | Thread or standalone tweet |
-| **YouTube** | High — visual narrative | 5-15 facts | Animated graphs, B-roll w/ data | Video essay / teardown |
-| **Blog / article** | Maximum | 10+ facts | Inline graphs, tables, interactives| Longread |
-| **Email** | Medium — 1 core insight | 2-3 facts | Big number, mini-infographic | Digest / newsletter |
-
-## Protocol
-
-### Step 0 — Data gathering and validation
-1. Procure input data: facts, sources, topic, persona, platform.
-2. Undergo **data validation** via the 6 criteria.
-3. Weed out unreliable data — work exclusively with verified figures.
-
-### Step 1 — Determining the data story type
-1. Analyze the character of the data — trend? comparison? anomaly?
-2. Select the type (or a combo of 2 types).
-3. Select the narrative structure fitting that type.
-
-### Step 2 — The Core Insight
-1. From the pile of data — abstract **one core insight** (the main idea).
-2. Formulate it within a single sentence.
-3. Check: is the insight **non-obvious**? If it's common knowledge — it's not an insight.
-
-### Step 3 — Hook-Fact
-1. Select the **strongest / most unexpected fact** from the data.
-2. This serves as the headline, hook, or opening sentence.
-3. Check: does the fact evoke a "really?!" or "I had no idea" reaction?
-
-### Step 4 — Constructing the narrative
-1. Distribute the data across the chosen type's narrative structure.
-2. Apply **1-3 persuasion techniques**.
-3. For each fact — spell out "what it means" (not merely a digit, but an interpretation).
-4. Conclusion — **takeaway + CTA** (what to do with this newly acquired knowledge).
-
-### Step 5 — Visualization
-1. Decide if visualization is even requisite (skip if only 1-3 numbers).
-2. Select visualization type based on the mapping table.
-3. Describe the visualization (for `$visual-brief` or the designer).
-
-### Step 6 — Platform adaptation
-1. Ascertain depth and number of facts from the platform table.
-2. Adapt the format (longread vs carousel vs thread).
-3. Adapt the voice/diction to the target persona.
-
-### Step 7 — Final Review
-Run through the filter:
-1. Does the story make sense **without numbers**? (numbers should confirm, not supplant the story).
-2. Do the numbers make sense **without context**? If not — apply scaling.
-3. Is the insight **non-obvious**?
-4. Is the data **free of misleading elements**?
-
-## Validation (Quality Gate)
-
-A data story is ready if:
-
-- [ ] All data passed validation (6 criteria)
-- [ ] A data story type is designated
-- [ ] The core insight is formulated (1 sentence, non-obvious)
-- [ ] A Hook-fact is appointed (evokes a reaction)
-- [ ] The narrative mimics the targeted type's structure
-- [ ] 1-3 persuasion techniques are harnessed
-- [ ] Each fact carries an interpretation ("what it means")
-- [ ] Visualization chosen (or lack thereof justified)
-- [ ] Adapted for platform and persona
-- [ ] Final review cleared (4 questions)
-
-## Handoff
-
-| Consumer | How it's used |
-|-------------|---------------|
-| `$content-brief` | Data story serves as the core of a specific content piece |
-| `$social-media-formats` | Adapting a data story into social media formats |
-| `$visual-brief` | Outlines visualization specs for the designer |
-| `$seo-copywriting` | Data story anchors an SEO-driven article |
-| `$email-copywriting` | Hook-fact + insight embedded into an email |
-| Copywriter (agent) | Narrative framework guides the draft |
-
-## Integration with other skills
-
-| Skill | Interaction |
-|------|----------------|
-| `$topic-research` | **Source:** data and facts pertinent to the topic |
-| `$source-verification` | **Source:** verified statistical data |
-| `$audience-analysis` | **Source:** persona → dictates depth and tone |
-| `$platform-strategy` | **Source:** platform → dictates volume and format |
-| `$fact-checking` | **Consumer:** verifies data assertions in the final draft |
-| `$visual-brief` | **Consumer:** receives spec for visualizations |
-
-## Anti-patterns
-
-| Error | Why it's bad | How to do it right |
-|--------|-------------|---------------|
-| Contextless numbers | "A $4.2T market" — so what? Is that a lot? | Scaling: "That's bigger than Germany's GDP" |
-| Data eclipsing story | A laundry list of facts sans narrative | Structure first, data as confirmation second |
-| Cherry-picking | Showcasing only the "pretty" data | Display the full picture, "ugly" facts included |
-| Correlation = Causation | "Ice cream → drownings" | Explicitly note: correlation, not causation |
-| Fossilized data | "A 2019 study" in 2026 | Keep data ≤ 2 years old (≤ 1 year for fast-moving markets) |
-| Fact spam | 15 facts crammed into one post | 2-5 facts for social; save the rest for longreads |
-| No insight | Everyone already knew it | Insight = a non-obvious deduction drawn from data |
-| Fabricated data | The agent hallucinates "statistics" | Utilize strictly verified sources |
-
-## Output Template
-
-```
-### Data Story — [Topic]
-
-**Type:** [Trend / Comparison / Anomaly / Correlation / Forecast / Scale]
-**Persona:** [from $audience-analysis]
-**Platform:** [platform]
-**Data sources:** [list w/ years]
-
----
-
-#### Core Insight
-[One sentence — the central, non-obvious deduction derived from the data]
-
-#### Hook-Fact (Headline material)
-[The most potent / surprising fact — evocative of a reaction]
-
----
-
-#### Narrative
-
-**Context:** [Why this matters right now — 1-2 sentences]
-
-**[Structure dictated by data story type]:**
-
-| Fact | Source | What it means |
-|------|----------|---------------|
-| [Fact 1 — specific number] | [source, year] | [interpretation tailored to persona] |
-| [Fact 2 — specific number] | [source, year] | [interpretation tailored to persona] |
-| [Fact 3 — specific number] | [source, year] | [interpretation tailored to persona] |
-
-**Insight:** [Non-obvious deduction — the connective tissue between the isolated numbers]
-
-**Takeaway / CTA:** [What to do with this info → tied to the content's objective]
-
----
-
-#### Persuasion Techniques
-1. [Technique 1] — application: "[excerpt from text]"
-2. [Technique 2] — application: "[excerpt from text]"
-
-#### Visualization
-
-| Element | Visual type | Data | Platform specs |
-|---------|------------|--------|---------------------|
-| [what's being visualized] | [line chart / bar / big number / ...] | [what data] | [dimensions, style] |
-
-> N/A — if data is scant (1-3 facts) and visualization is unwarranted.
-
-#### Data Validation
-
-| Criterion | Status | Note |
-|----------|--------|------------|
-| Recency (≤ 2 years) | ✅/❌ | [data year] |
-| Reliable source | ✅/❌ | [who conducted it] |
-| Sufficient sample | ✅/❌ | [N respondents/points] |
-| Not out of context | ✅/❌ | |
-| Causation vs Correlation | ✅/❌ | |
-| Complete (no cherry-picking) | ✅/❌ | |
-
----
-
-#### Platform Adaptation
-- **Depth:** [high / medium / low]
-- **Number of facts:** [N]
-- **Format:** [carousel / post / thread / longread / video]
-- **Language:** [adaptations specific to persona]
-
-**→ Next step:** forward to `$content-brief` / `$visual-brief` / Copywriter
-```
+---
+name: data-storytelling
+description: Turning data into narratives — data story types, persuasion techniques, data-driven visualization, platform adaptation
+type: triggered
+domain: content
+owners:
+  - copywriter
+gates:
+  - COPYWRITER
+tech: []
+topic: []
+triggers:
+  - "data-storytelling"
+  - "Data-сторителлинг"
+  - "Превращение данных в нарративы — типы data stories, техники убедительности, визуализация по типу данных, платформенная адаптация"
+related: []
+budget_lines: 309
+schema_version: 1
+---
+# Data Storytelling — Data-Driven Narratives
+
+## When to use
+- After **gathering facts and statistics** (`$topic-research`, `$source-verification`) — to structure them into a story.
+- When creating **analytical content** — articles, carousels, infographics, reports.
+- When **justifying a position** — case studies, comparisons, predictions.
+- To **reinforce any content** with data — even 1-2 numbers make a post more persuasive.
+
+## Input
+
+| Field | Required | Description |
+|------|:-----------:|----------|
+| Data / facts | ✅ | Numbers, statistics, research findings — the building blocks of the story |
+| Data sources | ✅ | Where the data comes from: research, analytics, API, report, `$source-verification` |
+| Topic / context | ✅ | What we are talking about — mapping to the content topic |
+| Target persona | ✅ | From `$audience-analysis` — dictates depth and language |
+| Platform | ✅ | Determines volume, visualization format, and depth |
+| Content goal | ⬚ | To inform / persuade / engage / convert |
+| Data story type | ⬚ | If omitted — the agent determines it based on the data's character |
+
+> If data is absent — **ask the user** or redirect to `$topic-research` / `$source-verification`. Do not fabricate numbers.
+
+## Data Story Types
+
+| Type | When to use | Question it answers | Example |
+|-----|-------------------|----------------------------|--------|
+| **Trend** | Data shows change over time | "What's happening and where is it heading?" | "The market for AI agents grew by 340% in a year" |
+| **Comparison** | Comparing 2+ subjects or periods | "Who is better / what has changed?" | "Email vs Telegram: Marketing channel ROI" |
+| **Anomaly** | An unexpected fact, an outlier, a contradiction | "Why is this strange?" | "87% of companies use AI, but only 12% see ROI" |
+| **Correlation** | Connection between two factors | "What is linked to what?" | "Companies with a content strategy generate 6× more leads" |
+| **Forecast** | Data allows for extrapolation | "What happens next?" | "By 2027, 50% of content will be AI-generated" |
+| **Scale** | Showcasing the magnitude of a phenomenon | "Just how big/small is this?" | "7.5M blog posts are published daily — that's 87 posts a second" |
+
+> The type dictates the narrative structure and choice of visualization.
+
+## Narrative structures by type
+
+### Trend
+```
+Context (what was) → Inflection point → Trend (data) → What it means → Forecast/CTA
+```
+
+### Comparison
+```
+Broad context → Subject A (data) → Subject B (data) → Key differentiator → Conclusion
+```
+
+### Anomaly
+```
+Expectation (what everyone thinks) → Reality (data) → Why it's the case → Insight → Conclusion
+```
+
+### Correlation
+```
+Observation → Connection (data A + data B) → Cause/mechanism → What to do about it
+```
+
+### Forecast
+```
+Current state (data) → Growth/decline factors → Forecast → What to do right now
+```
+
+### Scale
+```
+Abstract number → Scaling (translating to relatable terms) → Why it matters → Conclusion
+```
+
+## Persuasion techniques
+
+| Technique | Description | Example |
+|---------|----------|--------|
+| **Scaling** | Translating an abstract number into something relatable | "$3.5 trillion — that's more than the UK's GDP" |
+| **Simile / Comparison** | Juxtaposing with something familiar | "It's as if every third resident of London..." |
+| **Antithesis** | Contrasting expectation and reality | "Everyone thinks X. In reality — Y" |
+| **Metaphor** | Converting data into imagery | "Email marketing is like fishing: 95% of the catch depends on the bait (subject line)" |
+| **Personalization** | Bringing it down to a single person's level | "The average marketer spends 6 hours a week planning content" |
+| **Before/After** | Showing the state before and after | "Before automation: 3 days. After: 15 minutes" |
+| **Rule of Three** | Three facts are remembered best | "Faster. Cheaper. More efficient." |
+| **Q&A** | Asking a question, answering with data | "How much does a lead cost? $5 on Instagram vs $50 on LinkedIn" |
+
+> Use 1-3 techniques per data story. Any more results in overload.
+
+## Visualization by data type
+
+| Data / Story type | Recommended visualization | When NOT to use |
+|--------------------|--------------------------|-----------------------|
+| **Trend** (time) | Line chart, Area chart | Pie chart (lacks a time axis) |
+| **Comparison** (categories) | Bar chart (horizontal), Grouped bar | Line chart (implies continuity) |
+| **Proportions** (parts of whole)| Pie chart (≤ 5 segments), Stacked bar | Pie chart with 10+ segments |
+| **Correlation** (two vars) | Scatter plot, Bubble chart | Bar chart (doesn't show connection) |
+| **Scale** (one big fact) | Big number + context, Pictogram | Complex graph (kills the impact) |
+| **Ranking** (order) | Horizontal bar (sorted) | Pie chart (no inherent order) |
+| **Process** (stages) | Funnel chart, Sankey diagram | Pie chart |
+| **For social media** | Carousel w/ 1 graph per slide, infographics | Complex dashboards |
+
+> If data is sparse (1-3 numbers) — visualization is unnecessary. A prominent number within the text suffices.
+
+## Data Validation (mandatory)
+
+Prior to building a narrative — verify the data:
+
+| Criterion | Question | Risk if unverified |
+|----------|--------|----------------------|
+| **Recency** | When was the data gathered? | Outdated statistics = false conclusions |
+| **Source** | Who conducted the research? Is there a methodology? | Unreliable source = zero trust |
+| **Sample size** | How many respondents / data points? | Small sample = mere coincidence |
+| **Context** | Is the data cherry-picked? | Cherry-picking = manipulation |
+| **Causation vs Correlation**| Is the link causal or coincidental? | False causality = erroneous takeaway |
+| **Completeness** | Are both "pro" AND "con" data shown? | One-sidedness = bias |
+
+> If data fails validation — discard it. Fewer, yet reliable numbers are vastly superior.
+
+## Platform adaptation
+
+| Platform | Depth | Number of facts | Visualization | Format |
+|-----------|---------|-------------------|-------------|--------|
+| **LinkedIn** | High — comprehensive analysis | 5-10 facts | Infographics, carousel, graph screenshot | Long post / article |
+| **Instagram** | Medium — 1 insight + 2-3 facts | 2-4 facts | Carousel (1 fact = 1 slide), infographics | Carousel / reel |
+| **Telegram** | Medium-to-High | 3-7 facts | Embedded tables, link to full report | Well-formatted post |
+| **Twitter/X** | Low — 1 hook-fact | 1-2 facts | Graph screenshot / big number | Thread or standalone tweet |
+| **YouTube** | High — visual narrative | 5-15 facts | Animated graphs, B-roll w/ data | Video essay / teardown |
+| **Blog / article** | Maximum | 10+ facts | Inline graphs, tables, interactives| Longread |
+| **Email** | Medium — 1 core insight | 2-3 facts | Big number, mini-infographic | Digest / newsletter |
+
+## Protocol
+
+### Step 0 — Data gathering and validation
+1. Procure input data: facts, sources, topic, persona, platform.
+2. Undergo **data validation** via the 6 criteria.
+3. Weed out unreliable data — work exclusively with verified figures.
+
+### Step 1 — Determining the data story type
+1. Analyze the character of the data — trend? comparison? anomaly?
+2. Select the type (or a combo of 2 types).
+3. Select the narrative structure fitting that type.
+
+### Step 2 — The Core Insight
+1. From the pile of data — abstract **one core insight** (the main idea).
+2. Formulate it within a single sentence.
+3. Check: is the insight **non-obvious**? If it's common knowledge — it's not an insight.
+
+### Step 3 — Hook-Fact
+1. Select the **strongest / most unexpected fact** from the data.
+2. This serves as the headline, hook, or opening sentence.
+3. Check: does the fact evoke a "really?!" or "I had no idea" reaction?
+
+### Step 4 — Constructing the narrative
+1. Distribute the data across the chosen type's narrative structure.
+2. Apply **1-3 persuasion techniques**.
+3. For each fact — spell out "what it means" (not merely a digit, but an interpretation).
+4. Conclusion — **takeaway + CTA** (what to do with this newly acquired knowledge).
+
+### Step 5 — Visualization
+1. Decide if visualization is even requisite (skip if only 1-3 numbers).
+2. Select visualization type based on the mapping table.
+3. Describe the visualization (for `$visual-brief` or the designer).
+
+### Step 6 — Platform adaptation
+1. Ascertain depth and number of facts from the platform table.
+2. Adapt the format (longread vs carousel vs thread).
+3. Adapt the voice/diction to the target persona.
+
+### Step 7 — Final Review
+Run through the filter:
+1. Does the story make sense **without numbers**? (numbers should confirm, not supplant the story).
+2. Do the numbers make sense **without context**? If not — apply scaling.
+3. Is the insight **non-obvious**?
+4. Is the data **free of misleading elements**?
+
+## Validation (Quality Gate)
+
+A data story is ready if:
+
+- [ ] All data passed validation (6 criteria)
+- [ ] A data story type is designated
+- [ ] The core insight is formulated (1 sentence, non-obvious)
+- [ ] A Hook-fact is appointed (evokes a reaction)
+- [ ] The narrative mimics the targeted type's structure
+- [ ] 1-3 persuasion techniques are harnessed
+- [ ] Each fact carries an interpretation ("what it means")
+- [ ] Visualization chosen (or lack thereof justified)
+- [ ] Adapted for platform and persona
+- [ ] Final review cleared (4 questions)
+
+## Handoff
+
+| Consumer | How it's used |
+|-------------|---------------|
+| `$content-brief` | Data story serves as the core of a specific content piece |
+| `$social-media-formats` | Adapting a data story into social media formats |
+| `$visual-brief` | Outlines visualization specs for the designer |
+| `$seo-copywriting` | Data story anchors an SEO-driven article |
+| `$email-copywriting` | Hook-fact + insight embedded into an email |
+| Copywriter (agent) | Narrative framework guides the draft |
+
+## Integration with other skills
+
+| Skill | Interaction |
+|------|----------------|
+| `$topic-research` | **Source:** data and facts pertinent to the topic |
+| `$source-verification` | **Source:** verified statistical data |
+| `$audience-analysis` | **Source:** persona → dictates depth and tone |
+| `$platform-strategy` | **Source:** platform → dictates volume and format |
+| `$fact-checking` | **Consumer:** verifies data assertions in the final draft |
+| `$visual-brief` | **Consumer:** receives spec for visualizations |
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|--------|-------------|---------------|
+| Contextless numbers | "A $4.2T market" — so what? Is that a lot? | Scaling: "That's bigger than Germany's GDP" |
+| Data eclipsing story | A laundry list of facts sans narrative | Structure first, data as confirmation second |
+| Cherry-picking | Showcasing only the "pretty" data | Display the full picture, "ugly" facts included |
+| Correlation = Causation | "Ice cream → drownings" | Explicitly note: correlation, not causation |
+| Fossilized data | "A 2019 study" in 2026 | Keep data ≤ 2 years old (≤ 1 year for fast-moving markets) |
+| Fact spam | 15 facts crammed into one post | 2-5 facts for social; save the rest for longreads |
+| No insight | Everyone already knew it | Insight = a non-obvious deduction drawn from data |
+| Fabricated data | The agent hallucinates "statistics" | Utilize strictly verified sources |
+
+## Output Template
+
+```
+### Data Story — [Topic]
+
+**Type:** [Trend / Comparison / Anomaly / Correlation / Forecast / Scale]
+**Persona:** [from $audience-analysis]
+**Platform:** [platform]
+**Data sources:** [list w/ years]
+
+---
+
+#### Core Insight
+[One sentence — the central, non-obvious deduction derived from the data]
+
+#### Hook-Fact (Headline material)
+[The most potent / surprising fact — evocative of a reaction]
+
+---
+
+#### Narrative
+
+**Context:** [Why this matters right now — 1-2 sentences]
+
+**[Structure dictated by data story type]:**
+
+| Fact | Source | What it means |
+|------|----------|---------------|
+| [Fact 1 — specific number] | [source, year] | [interpretation tailored to persona] |
+| [Fact 2 — specific number] | [source, year] | [interpretation tailored to persona] |
+| [Fact 3 — specific number] | [source, year] | [interpretation tailored to persona] |
+
+**Insight:** [Non-obvious deduction — the connective tissue between the isolated numbers]
+
+**Takeaway / CTA:** [What to do with this info → tied to the content's objective]
+
+---
+
+#### Persuasion Techniques
+1. [Technique 1] — application: "[excerpt from text]"
+2. [Technique 2] — application: "[excerpt from text]"
+
+#### Visualization
+
+| Element | Visual type | Data | Platform specs |
+|---------|------------|--------|---------------------|
+| [what's being visualized] | [line chart / bar / big number / ...] | [what data] | [dimensions, style] |
+
+> N/A — if data is scant (1-3 facts) and visualization is unwarranted.
+
+#### Data Validation
+
+| Criterion | Status | Note |
+|----------|--------|------------|
+| Recency (≤ 2 years) | ✅/❌ | [data year] |
+| Reliable source | ✅/❌ | [who conducted it] |
+| Sufficient sample | ✅/❌ | [N respondents/points] |
+| Not out of context | ✅/❌ | |
+| Causation vs Correlation | ✅/❌ | |
+| Complete (no cherry-picking) | ✅/❌ | |
+
+---
+
+#### Platform Adaptation
+- **Depth:** [high / medium / low]
+- **Number of facts:** [N]
+- **Format:** [carousel / post / thread / longread / video]
+- **Language:** [adaptations specific to persona]
+
+**→ Next step:** forward to `$content-brief` / `$visual-brief` / Copywriter
+```