AgentSkillsCN

Story Craft

故事创作

SKILL.md

Story Craft - Empathy Ledger Story Generation

Transform raw transcripts into authentic, culturally-sensitive stories with the Empathy Ledger voice and craft.

Purpose

This skill helps you:

  1. Identify poor-quality stories (raw transcripts, AI-generated, no narrative)
  2. Generate authentic stories from transcripts with proper narrative structure
  3. Apply Empathy Ledger style and brand to storytelling
  4. Ensure cultural sensitivity and respect in every story

Empathy Ledger Story Principles

The Voice

Authentic & Grounded:

  • First-person or close third-person perspective
  • Natural, conversational tone
  • Real experiences, not abstractions
  • Specific details that bring moments to life

Culturally Respectful:

  • Honor Indigenous voices and perspectives
  • Acknowledge Country and cultural background
  • Use culturally appropriate language
  • Respect sacred and sensitive content

Human-Centered:

  • Focus on people, not systems
  • Show vulnerability and strength
  • Capture emotional truth
  • Celebrate resilience and wisdom

Story Structure

Every Empathy Ledger story should have:

  1. Opening Hook - A moment, scene, or question that draws you in
  2. Context - Who, where, when (cultural background, place, time)
  3. Journey - The narrative arc (challenge, experience, transformation)
  4. Insight - What was learned, felt, or understood
  5. Resonance - Why this matters beyond the individual

What Good Looks Like

✅ GOOD Story Example:

code
Title: Coming Home to Country

The first time I returned to Palm Island after the cyclone, I couldn't
recognize the shoreline. The mangroves—those ancient sentinels that had
stood for generations—were stripped bare, their branches reaching skyward
like desperate prayers.

My grandmother used to say that Country remembers everything. As I walked
through what remained of our community, I understood what she meant. The
land held the weight of our stories, even when the buildings didn't.

We gathered at the old meeting place, the one that survived somehow. Aunty
Rose brought tea in a chipped thermos. Uncle James had his guitar. And
slowly, story by story, we began to rebuild—not just the structures, but
the connections that make us who we are.

The cyclone took our houses. But it couldn't touch what binds us to this
place, to each other, to the generations who came before and those who'll
come after.

Why it works:

  • Opens with a vivid, specific image
  • Grounded in place and cultural context
  • Personal voice with emotional truth
  • Narrative arc (return → recognition → gathering → insight)
  • Respects Elders and cultural knowledge
  • Universal resonance about resilience

❌ BAD Story Example (Raw Transcript):

code
Speaker 1: [00:00:15] So yeah, um, we came back after the cyclone and,
you know, like, everything was just, um, gone basically.

Interviewer: [00:00:23] Can you describe what you saw?

Speaker 1: [00:00:25] Well, like I said, um, the mangroves were all, you
know, destroyed and stuff. And yeah, it was pretty bad. My grandmother
always talked about, um, you know, the land and how it remembers things
[inaudible].

Interviewer: [00:00:45] What happened next?

Speaker 1: [00:00:47] We, uh, we all got together at the meeting place.
Aunty Rose was there with tea and Uncle James had his guitar, you know.
And we just, like, started telling stories and stuff.

Why it fails:

  • Timecodes and speaker labels (transcript artifact)
  • No narrative structure or flow
  • Filler words ("um," "like," "you know")
  • Interviewer questions break immersion
  • No emotional depth or specificity
  • Reads like raw data, not a story

Quality Markers to Check

RED FLAGS (Poor Quality):

  • ❌ Contains timecodes [00:00:00]
  • ❌ Speaker labels (Speaker 1:, Interviewer:)
  • ❌ Transcript artifacts ([inaudible], [crosstalk], um, uh)
  • ❌ Question-answer format (interview structure visible)
  • ❌ No paragraphs (wall of text)
  • ❌ Generic AI language ("As an AI...", "I cannot...")
  • ❌ No cultural context or grounding
  • ❌ Abstract, not specific
  • ❌ No emotional resonance
  • ❌ Feels like a report, not a story

GREEN LIGHTS (Good Quality):

  • ✅ Clear narrative voice (first or third person)
  • ✅ Paragraph structure with natural breaks
  • ✅ Specific, vivid details
  • ✅ Cultural context acknowledged
  • ✅ Emotional truth and authenticity
  • ✅ Clear arc (beginning, middle, end)
  • ✅ Respectful language and framing
  • ✅ Stands alone (no interviewer needed)
  • ✅ Reads like a story, not data
  • ✅ Human warmth and connection

Transformation Process

Step 1: Analyze the Transcript

Look for:

  • Key moments - What actually happened?
  • Emotional beats - How did they feel?
  • Cultural context - Where, when, who?
  • Themes - What's the deeper meaning?
  • Unique details - What makes this specific?

Step 2: Identify the Story Arc

Every story needs:

  • Inciting moment - What changed or started things?
  • Journey - What happened along the way?
  • Transformation - What shifted or was learned?
  • Resolution - Where did it land?

Step 3: Craft the Narrative

Opening:

  • Start with a moment, not background
  • Use specific, sensory details
  • Establish voice immediately

Middle:

  • Show, don't tell
  • Include dialogue if natural
  • Build emotional tension
  • Honor the speaker's perspective

Closing:

  • Land on insight or resonance
  • Connect to something larger
  • Leave room for reflection

Step 4: Cultural Sensitivity Review

Check:

  • Is the cultural background respectfully presented?
  • Are Elders and knowledge holders honored?
  • Is the language appropriate and respectful?
  • Are sacred or sensitive topics handled appropriately?
  • Does it avoid stereotypes or appropriation?

Usage

Identify Poor-Quality Story

code
Task: Review this story for quality issues:

[paste story content]

Analyze for:
- Transcript artifacts (timecodes, speaker labels)
- Lack of narrative structure
- Cultural sensitivity concerns
- Missing Empathy Ledger voice

Transform Transcript to Story

code
Task: Transform this transcript into an authentic Empathy Ledger story:

Transcript:
[paste raw transcript]

Cultural Context:
- Storyteller: [name, cultural background]
- Location: [place, Country]
- Topic: [what it's about]
- Themes: [key themes]

Create a story that:
1. Removes all transcript artifacts
2. Establishes clear narrative voice
3. Honors cultural context
4. Follows Empathy Ledger principles
5. Has emotional authenticity and resonance

Batch Quality Audit

code
Task: Audit these stories and flag those needing transformation:

[list of story IDs or titles]

For each, identify:
- Quality score (1-10)
- Specific issues
- Recommendation (keep, transform, delete)

Anti-Patterns to Avoid

Don't:

  • Don't just remove timecodes and call it done
  • Don't make it overly formal or academic
  • Don't add flowery language that wasn't there
  • Don't erase the speaker's voice or personality
  • Don't appropriate or misrepresent cultural content
  • Don't turn it into a report or summary
  • Don't add details that weren't in the source
  • Don't lose the emotional truth

Do:

  • Do preserve the speaker's authentic voice
  • Do structure for clarity and flow
  • Do honor the emotional truth
  • Do respect cultural context
  • Do make it readable and engaging
  • Do keep specific, real details
  • Do ensure cultural sensitivity
  • Do create something that resonates

Examples by Story Type

Personal Experience Story

Before (Transcript):

code
Speaker 12: [00:15:30] So, um, when I was a kid, you know, we used to go
fishing with my grandfather on the weekends. And, like, he would tell us
stories about, um, the old ways and stuff. [pause] Yeah, it was really
special, you know?

After (Story):

code
Every Saturday morning, my grandfather would wake us before dawn. "The fish
don't wait for sleepyheads," he'd say, but we knew it wasn't really about
the fish.

On those quiet boat rides, between casts, he'd share stories—about his
grandfather, about Country, about the ways things used to be. His voice
would get softer when he talked about the old ways, like he was sharing
something sacred.

I didn't always understand what he was teaching us then. But now, when I
take my own kids fishing, I hear his words in mine.

Community Event Story

Before (Transcript):

code
Interviewer: Can you describe the cultural festival?

Speaker 3: [00:05:12] Yeah, so we had like about 200 people show up. There
was traditional dancing, um, food stalls, and Uncle Robert did a Welcome to
Country. It was, you know, really good to see everyone come together like
that.

After (Story):

code
When Uncle Robert's voice carried across the gathering—deep and steady,
acknowledging Country and Ancestors—two hundred people fell silent. Some
of us had tears in our eyes.

The dancers' feet struck the earth in rhythms that pre-date memory. Kids
watched, wide-eyed, as their grandparents moved with a grace that defied
age. The smell of traditional foods drifted through the crowd.

This wasn't just a festival. This was our community saying: we're still
here, still strong, still connected to who we are and where we come from.

Quality Standards

A well-crafted Empathy Ledger story should:

  1. Pass the read-aloud test - Sound natural when spoken
  2. Stand alone - Make sense without transcript or interview
  3. Honor the source - Keep the truth and voice intact
  4. Engage emotionally - Make readers feel something
  5. Respect culture - Handle cultural content appropriately
  6. Flow naturally - Have clear structure and pacing
  7. Resonate beyond - Connect to something universal

When to Use This Skill

Invoke when:

  • Reviewing story quality in the admin panel
  • Transforming transcripts into stories
  • Auditing batches of stories for issues
  • Setting up automated quality checks
  • Training on Empathy Ledger storytelling standards
  • Reviewing cultural sensitivity of content
  • Preparing stories for publication or sharing

Related Skills

  • review-cultural - For cultural sensitivity review
  • data-analysis - For extracting themes from transcripts
  • design-component - For storyteller card display

Remember: Every story in Empathy Ledger represents a real person sharing their truth. Our job is to honor that truth with craft, respect, and authenticity.