narrative-framing

$npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/narrative-framing

Constructs theory-driven narratives by guiding evidence-based reading

  • Solves complex literature analysis by building structured arguments from themes
  • Uses paper-search, paper-research, thematic-coding, and web-research tools
  • Decides based on intellectual judgment and pattern recognition across sources
  • Delivers a coherent, evidence-supported argument with counter-evidence synthesis

SKILL.md

.github/skills/narrative-framingView on GitHub ↗
---
name: narrative-framing
description: Theory-driven reading tactic — define a theoretical framework first, then guide reading to fill it with evidence. Five stages (theme identification, argument construction, evidence collection, counter-evidence, synthesis). The most intellectually demanding tactic.
execution: tactic
used-by: literature-survey
---

# Narrative Framing

Define a theoretical framework first, then guide reading to fill it with evidence.

## Stages

1. **Theme identification** — from initial reading, identify 3-5 major themes
2. **Argument construction** — define the narrative arc (thesis → evidence → synthesis)
3. **Evidence collection** — targeted reading to support each theme
4. **Counter-evidence** — actively search for work that challenges the narrative
5. **Synthesis** — weave themes into coherent argument

## Available SOPs

- `paper-search` (import) — targeted reading for evidence
- `paper-research` (import) — deep reading of key supporting/opposing papers
- `thematic-coding` (subagent) — identify patterns across papers
- `web-research` (import) — blogs, opinion pieces, industry perspectives

## Execution Guidance

- This is the most "intellectual" tactic — requires judgment about what matters
- thematic-coding helps identify patterns across papers
- CC must actively seek counter-evidence (intellectual honesty)
- Output is not a list but a structured argument
- Themes should emerge from reading, not be imposed a priori

More from yogsoth-ai/de-anthropocentric-research-engine