Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Apply narrative research methods to understand human experience through stories, analyzing narrative structure, temporality, and meaning-making in life stories and oral histories. Use this skill when the user needs to analyze how people construct meaning through storytelling, examine narrative structure and plot, conduct life story or oral history research, or when they ask 'how do stories shape identity', 'how do I analyze a life narrative', or 'what does this story reveal about experience'.
.claude/skills/asgard-ai-platform-grad-narrative/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 106% | 0% |
Narrative research studies human experience through the stories people tell about their lives. Stories are not merely reports of events but active constructions of meaning — organizing experience temporally, assigning causality, and shaping identity. The methodology preserves the wholeness of narratives rather than fragmenting them, attending to structure (how the story is told), content (what is told), and context (why and to whom).
IRON LAW: In narrative research, the STORY is the unit of analysis —
fragmenting narratives into coded themes destroys the temporal and
contextual meaning. If you reduce stories to thematic codes, you are
doing thematic analysis, NOT narrative research.Key assumptions:
Conduct narrative interviews using a single generative question (e.g., "Tell me the story of..."). Allow the narrator to structure the telling. Avoid interrupting with probes until the main narrative is complete. Collect life stories, oral histories, or episode-specific narratives depending on the research focus.
Examine HOW the story is told using structural elements:
| Element | Description | |---------|-------------| | Abstract | What is this story about? | | Orientation | Who, when, where, what situation? | | Complicating action | Then what happened? (the plot) | | Evaluation | So what? Why does this matter to the narrator? | | Resolution | What finally happened? | | Coda | Return to the present; moral or lesson |
(Labov's structural model; adapt as needed for non-Western narrative forms.)
Analyze WHAT the story conveys: turning points, character positioning, agency, causality, and moral framing. Examine how the narrator positions themselves (hero, victim, survivor, agent). Identify master narratives the story aligns with or resists.
Situate the narrative within broader social, cultural, and historical contexts. Present findings as re-storied narratives (Clandinin and Connelly) or analytic narratives that preserve the story's integrity while offering scholarly interpretation.
markdown## Narrative Analysis: [Context] ### Narrator Profile - Pseudonym: [name] - Context: [relevant background] - Narrative type: [life story / episodic / oral history] ### Narrative Structure | Structural Element | Content | |-------------------|---------| | Abstract | [summary of what the story is about] | | Orientation | [setting, characters, initial situation] | | Complicating action | [key events and turning points] | | Evaluation | [narrator's assessment of meaning] | | Resolution | [how events resolved] | | Coda | [return to present, lesson drawn] | ### Identity Positioning - Self-positioning: [how the narrator presents themselves] - Agency: [active agent / constrained / victim / survivor] - Master narratives: [aligned with or resisting which cultural stories] ### Temporal and Contextual Meaning - Turning points: [pivotal moments that reorganize the narrative] - Causality: [how the narrator explains why things happened] - Silences: [what is notably absent from the story] ### Implications 1. [What this narrative reveals about the phenomenon] 2. [How it connects to broader social or cultural narratives]
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 16,206 | 8,276 | -49% | 1 | 1 | 0% | 2,794 | 2,411 | -14% | 0 | 0 | — |
case-01 | fail→pass | 32,049 | 21,210 | -34% | 1 | 1 | 0% | 4,560 | 4,285 | -6% | 0 | 0 | — |
case-02 | fail→pass | 33,515 | 15,969 | -52% | 1 | 1 | 0% | 4,776 | 3,455 | -28% | 0 | 0 | — |
case-03 | fail→pass | 33,454 | 20,988 | -37% | 1 | 1 | 0% | 4,822 | 4,611 | -4% | 0 | 0 | — |
case-04 | pass→pass | 12,324 | 10,454 | -15% | 1 | 1 | 0% | 2,031 | 2,777 | +37% | 0 | 0 | — |
case-22 | fail→fail | 21,366 | 12,771 | -40% | 1 | 1 | 0% | 3,180 | 3,374 | +6% | 0 | 0 | — |
case-05 | pass→pass | 16,049 | 10,035 | -37% | 1 | 1 | 0% | 2,266 | 2,556 | +13% | 0 | 0 | — |
case-06 | pass→pass | 5,312 | 4,541 | -15% | 1 | 1 | 0% | 938 | 1,985 | +112% | 0 | 0 | — |
case-07 | pass→pass | 7,721 | 9,024 | +17% | 1 | 1 | 0% | 1,118 | 2,484 | +122% | 0 | 0 | — |
case-08 | pass→pass | 7,801 | 7,303 | -6% | 1 | 1 | 0% | 1,112 | 2,283 | +105% | 0 | 0 | — |
case-09 | pass→pass | 10,165 | 4,574 | -55% | 1 | 1 | 0% | 1,523 | 1,935 | +27% | 0 | 0 | — |
case-10 | pass→pass | 14,502 | 10,497 | -28% | 1 | 1 | 0% | 2,339 | 2,752 | +18% | 0 | 0 | — |
case-11 | pass→pass | 12,453 | 12,117 | -3% | 1 | 1 | 0% | 1,865 | 2,907 | +56% | 0 | 0 | — |
case-12 | pass→pass | 13,006 | 10,743 | -17% | 1 | 1 | 0% | 2,123 | 2,880 | +36% | 0 | 0 | — |
case-13 | pass→pass | 11,263 | 10,509 | -7% | 1 | 1 | 0% | 1,791 | 2,884 | +61% | 0 | 0 | — |
case-14 | pass→pass | 18,695 | 14,938 | -20% | 1 | 1 | 0% | 2,401 | 3,427 | +43% | 0 | 0 | — |
case-15 | pass→pass | 11,574 | 9,800 | -15% | 1 | 1 | 0% | 1,685 | 2,466 | +46% | 0 | 0 | — |
case-16 | fail→pass | 11,719 | 7,769 | -34% | 1 | 1 | 0% | 1,577 | 2,358 | +50% | 0 | 0 | — |
case-17 | fail→pass | 17,049 | 38,497 | +126% | 1 | 1 | 0% | 1,231 | 2,535 | +106% | 0 | 0 | — |
case-18 | fail→pass | 15,307 | 8,876 | -42% | 1 | 1 | 0% | 2,459 | 2,756 | +12% | 0 | 0 | — |
case-19 | fail→pass | 12,652 | 10,596 | -16% | 1 | 1 | 0% | 1,979 | 2,775 | +40% | 0 | 0 | — |
case-20 | pass→pass | 15,814 | 15,100 | -5% | 1 | 1 | 0% | 2,539 | 3,867 | +52% | 0 | 0 | — |
case-23 | fail→pass | 16,617 | 8,409 | -49% | 1 | 1 | 0% | 2,709 | 2,536 | -6% | 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. 23 cases were attempted. The headline lift of +35 percentage points is the difference between those two pass rates over the 23 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.