
from social-data-analysis47
Guides theory-first, anomaly-driven qualitative analysis of interview data using Timmermans & Tavory's abductive methods.
This skill turns an agent into an expert qualitative research assistant for abductive analysis. It leads users through a structured, multi-phase workflow (theoretical preparation, familiarization and open coding, theoretical casing, anomaly analysis, memo writing, integration/testing, and write-up) to surface surprising observations and develop novel theoretical claims. The skill emphasizes iterative movement between data and theory, flagging anomalies and systematically testing alternative explanations.
Use this skill when you have interview transcripts or qualitative data and want a theory-first analysis that seeks explanatory surprises rather than simple summaries. It's appropriate for academic researchers, evaluators, or teams preparing article drafts who need help moving from coded excerpts to memo-driven theorizing.
phases/ directory (has_references=false in fetch output but the skill includes phase guides inline).{}
Abductive-analyst is a purely instructional skill with no executable scripts. It guides users through a 7-phase qualitative analysis process (Timmermans & Tavory's abductive method) for interview data. The SKILL.md is well-structured with clear phase descriptions, pause points, model recommendations, and references to external phase guides. No security concerns whatsoever — no code execution, network calls, or file system operations.
Clean, well-structured instructional skill. No security issues. Architecture is good with phase separation and external reference files, though frontmatter could include more metadata (tags, version). Usefulness is limited by the narrow academic niche.