Agent skill · research science · aiskillstore
reasoning-abductive
Generate and evaluate explanatory hypotheses from incomplete observations. Use when diagnosing anomalies, explaining unexpected outcomes, or inferring causes from effects. Produces ranked hypotheses with evidence and confidence scores.
Why this skill is useful
Provides a structured methodology for generating and evaluating hypotheses in diagnostic contexts, which the AI wouldn't reliably generate on its own.
What it needs
About 5k tokens when loaded. Last updated 2026-08-07. 405 stars on the source repository.
What this skill does
Abductive Reasoning Generate best explanations from observations. The logic of diagnosis and inference to cause. Type Signature When to Use Use abductive when: Anomaly or surprise present (metric deviation, unexpected behavior) Need to explain an observation (why did X happen?) Incomplete data, must infer cause Diagnostic context (errors, issues, failures) Multiple possible explanations exist Don't use when: Cause is already known (use causal execution instead) Need to compare decision alternatives Resolving stakeholder disagreements Four-Stage Process Stage 1: Observation Purpose: Transform raw data into structured anomaly description. Input: Process: 1. Quantify the deviation (% change, sigma from mean) 2. Identify temporal boundaries (when did it start?) 3. Scope the phenomenon (which segments affected?) 4. Rate surprise level (expected vs unexpected) Output: Stage 2: Hypotheses Generation Purpose: Generate diverse possible explanations without judgment. Rules: Generate ≥5 hypotheses (more is better initially) Include obvious AND non-obvious causes Consider internal AND external factors Don't filter yet - cast wide net Hypothesis Categories: Category Examples ---------- ---------- Technical Site issues, bugs, performance Product Features, pricing, positioning Market Competition, trends, seasonality Operational Team changes, process issues External Economy, regulations, events Output: Stage 3: Evidence Evaluation Purpose: Score each hypothesis against available evidence. For each hypothesis, evaluate: Criterion Question Score ----------- ---------- ------- Explanatory power Does it fully explain the anomaly? 0-1 Simplicity Fewest assumptions required? 0-1 Coherence Consistent with other known facts? 0-1 Testability Can we verify/falsify it? 0-1 Prior probability How likely independent of this data? 0-1 Evidence Collection: Scoring Formula: Stage 4: Best Explanation Purpose: Select most probable cause with confidence and next steps. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aiskillstore/reasoning-abductive