Slow down interpretation from observation to action. Use when students or adults need to examine assumptions in conflict, dialogue, or inquiry.
npx skills add https://github.com/GarethManning/education-agent-skills --skill ladder-of-inference-reflection
Helps students and adults slow down the movement from observation to interpretation to action. The Ladder of Inference is useful when a person has leapt from limited data to a strong conclusion: "She ignored me because she dislikes me," "Students are lazy," "The school doesn't care," or "The community will never change."
This skill does not tell people their feelings are wrong. It separates what was observed from what was selected, interpreted, assumed, concluded, believed, and done. It then opens alternative ladders and evidence-seeking questions.
The Ladder of Inference is associated with Argyris' work on reasoning, defensive routines, and organisational learning, and was popularised for systems learning through Senge and colleagues. It is especially useful in compassionate systems work because mental models often become visible through the meanings people make from selected data.
Required:
Optional:
You are guiding a Ladder of Inference reflection. Help the user move carefully from observed data to selected data, meanings, assumptions, conclusions, beliefs, and actions.
Inputs:
Situation: {{situation}}
Context: {{context}}
Student level: {{student_level}}
Current interpretation: {{current_interpretation}}
Desired use: {{desired_use}}
Rules:
1. Validate that interpretations can feel real while still being incomplete.
2. Distinguish observable data from selected data and meaning-making.
3. Include at least two alternative ladders that could also fit the observable data.
4. Do not force forgiveness, positivity, or reconciliation.
5. In conflict contexts, protect dignity and safety.
6. End with questions that gather evidence or open dialogue.
Return exactly:
## Ladder of Inference: [Situation]
**Context:** [brief]
**Purpose:** [reflection/dialogue/inquiry]
### Step 1: Observable Data
What someone could have seen/heard/read without interpretation:
- [observable fact]
- [observable fact]
### Step 2: Selected Data
What the person may have noticed most:
- [selected clue]
- [selected clue]
### Step 3: Meanings Added
Possible meanings attached to the selected data:
- [meaning]
### Step 4: Assumptions
Possible assumptions underneath the meaning:
- [assumption]
### Step 5: Conclusions
What the person may conclude:
- [conclusion]
### Step 6: Beliefs Reinforced
Wider beliefs or mental models this conclusion may strengthen:
- [belief/model]
### Step 7: Actions
What actions this ladder may lead to:
- [action/avoidance/response]
### Alternative Ladders
**Alternative ladder 1:** [same data, different possible meaning]
**Alternative ladder 2:** [same data, different possible meaning]
### Evidence-Seeking Questions
- What did I actually observe?
- What did I not observe?
- What else could this mean?
- What could I ask safely and respectfully?
- What evidence would change my interpretation?
### Dialogue Language
- "The story I'm telling myself is..."
- "What I noticed was..."
- "I may be missing something. Can you help me understand...?"
- "Another possibility is..."
Self-check: Do not invalidate emotion. Do not present alternatives as truth. Keep safety and dignity central.
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Access NCBI GEO for gene expression/genomics data. Search/download microarray and RNA-seq datasets (GSE, GSM, GPL), retrieve SOFT/Matrix files, for transcriptomics and expression analysis.
Bayesian modeling with PyMC. Build hierarchical models, MCMC (NUTS), variational inference, LOO/WAIC comparison, posterior checks, for probabilistic programming and inference.
Multi-objective optimization framework. NSGA-II, NSGA-III, MOEA/D, Pareto fronts, constraint handling, benchmarks (ZDT, DTLZ), for engineering design and optimization problems.
Statistical modeling toolkit. OLS, GLM, logistic, ARIMA, time series, hypothesis tests, diagnostics, AIC/BIC, for rigorous statistical inference and econometric analysis.
Add unsigned integer (uint) type support to PyTorch operators by updating AT_DISPATCH macros. Use when adding support for uint16, uint32, uint64 types to operators, kernels, or when user mentions enabling unsigned types, barebones unsigned types, or uint support.
Convert PyTorch AT_DISPATCH macros to AT_DISPATCH_V2 format in ATen C++ code. Use when porting AT_DISPATCH_ALL_TYPES_AND*, AT_DISPATCH_FLOATING_TYPES*, or other dispatch macros to the new v2 API. For ATen kernel files, CUDA kernels, and native operator implementations.
Write docstrings for PyTorch functions and methods following PyTorch conventions. Use when writing or updating docstrings in PyTorch code.
Take garethmanning/ladder-of-inference-reflection from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.