Legal Abductive Reasoning logo

Legal Abductive Reasoning

Community
THUYRan
legal-abductive-reasoning

Legal abductive reasoning skill for generating and evaluating the most reasonable explanatory hypotheses when evidence is incomplete or facts are ambiguous. Combines Mill’s five methods (agreement, difference, joint method, residues, concomitant variation) for structured causal inference. Triggered when the user needs to “propose possible legal explanations,” “generate defense/prosecution hypotheses,” “evaluate multiple readings of evidence,” or “infer the best explanation from incomplete facts.” Applies to case analysis, evidence evaluation, disputes over rule application, causal inference, and similar scenarios.

Overview

PublisherTHUYRan
RepositoryLegal-Skills-Chinese
Skill namelegal-abductive-reasoning
Stars
876
Forks
97
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by THUYRan on GitHub. Read the source before you install it.

Installation

Install the Legal Abductive Reasoning AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/THUYRan/Legal-Skills-Chinese/tree/main/English/skills/Legal-Abductive-Reasoning
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/THUYRan/Legal-Skills-Chinese.git /tmp/Legal-Skills-Chinese
mkdir -p .claude/skills
cp -r /tmp/Legal-Skills-Chinese/English/skills/Legal-Abductive-Reasoning .claude/skills/legal-abductive-reasoning
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Legal Abductive Reasoning in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Legal Abductive Reasoning on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Legal Abductive Reasoning is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Chinese source (authoritative): ../../skills/Legal-Abductive-Reasoning/SKILL.md

Core functions:

  • Fill explanatory gaps when evidence is incomplete
  • Construct plausible versions of the facts when facts are ambiguous
  • Infer implicit conditions when rules are unclear
  • Generate defeasible hypotheses

Legal Abductive Reasoning Skill

1. Core Concept: What Is Abductive Reasoning?

Abductive reasoning infers causes from results:

“Observe a phenomenon → propose the most reasonable explanation (hypothesis)”

Key features:

  • Defeasibility: Hypotheses can be overturned by new evidence (non-monotonic reasoning)
  • Creativity: Requires imagination and creative thinking
  • Explanatory focus: Concerns “why,” not merely “what”

Difference from deduction and induction:


2. Mill’s Five Methods: Tools for Hypothesis Generation and Testing

2.1 Method of Agreement — Find Necessary Conditions

Logic:

Case 1: Factors A, B, C → Phenomenon x occurs
Case 2: Factors A, D, E → Phenomenon x occurs
─────────────────────────────
∴ A is a necessary condition of x (no A → no x)

Legal application: Determine “necessary factual elements for a legal consequence”

Example: All theft cases involve “taking away” → “taking away” is a necessary condition of theft


2.2 Method of Difference — Find Sufficient Conditions

Logic:

Case 1: Factors A, B, C → Phenomenon x occurs
Case 2:     B, C   → Phenomenon x does not occur
─────────────────────────────
∴ A is a sufficient condition of x (A → x)

Legal application: Judge whether conduct is “sufficient” to cause a particular legal consequence

Example: Warning present → no accident; warning absent → accident → warning is a sufficient condition for preventing the accident


2.3 Joint Method of Agreement and Difference — Necessary + Sufficient

Logic: Apply agreement and difference together to strengthen causal inference

Legal application: Identify “key facts” or “decisive factors”


2.4 Method of Residues — Explain the Remainder

Logic:

Known: A+B+C → x+y+z
Known: B → y, C → z
────────────────────
∴ A → x (remainder explained by a new cause)

Legal application: Identify “additional factors” (e.g., discriminatory motive, mental harm)


2.5 Method of Concomitant Variation — Variable Relationships

Logic:

A increases → x increases
A decreases → x decreases
────────────────────
∴ A and x have a causal or common-cause relationship

Legal application: Identify degree relationships (e.g., degree of negligence and degree of harm)


3. Workflow: Five-Step Abductive Analysis

Step 1: Fact Identification

Goal: Clearly list known facts and identify information gaps

Operations:

  1. Distinguish proven facts from facts to be proven
  2. Label favorable and unfavorable facts
  3. Identify ambiguities and multiple possibilities

Output format:

markdown
## Known Facts
- [Fact 1]: [description] (evidence source / strength)
- [Fact 2]: [description] (evidence source / strength)

## Information Gaps
- [Gap 1]: [missing key information]

## Ambiguities
- [Ambiguity 1]: [fact open to multiple readings]

Step 2: Hypothesis Generation

Goal: Using Mill’s five methods, generate multiple competing explanations

Mandatory requirement: Generate at least 3 hypotheses, covering different angles

Hypothesis types:

  • Factual hypotheses: Different versions of what happened
  • Intentional hypotheses: Different explanations of motive and purpose
  • Legal hypotheses: Different judgments on whether elements are satisfied

Output format:

markdown
## Competing Hypotheses

### Hypothesis 1: [brief description]
- **Method based on**: [agreement / difference / joint / residues / concomitant variation]
- **Core claim**: [one-sentence summary]
- **Supporting evidence**: [list supporting facts]
- **Potential weaknesses**: [list unfavorable facts]
- **Defeasibility conditions**: [what evidence would overturn this hypothesis]

### Hypothesis 2: [brief description]
...

Step 3: Hypothesis Evaluation

Evaluation criteria (PEIRCE):

CriterionMeaningEvaluation question
Explanatory powerCoverageCan it explain all known facts?
SimplicityOccam’s razorDoes it introduce unnecessary entities?
TestabilityEmpirical possibilityCan it be verified or falsified?
Background fitConsistency with precedentIs it consistent with similar cases?

Evaluation matrix:

HypothesisExplanatory powerSimplicityTestabilityBackground fitOverall
H1High/Med/LowHigh/Med/LowHigh/Med/LowHigh/Med/Low⭐⭐⭐⭐⭐
H2...............

Step 4: Inference to Best Explanation

Decision:

  • Preferred hypothesis: Highest overall score
  • Alternative hypotheses: Near-scoring options retained
  • Key uncertainties: Information gaps that affect the judgment

Step 5: Validation & Refinement

Operations:

  1. List evidence that can be further collected
  2. Predict how new evidence would affect each hypothesis
  3. Set trigger conditions for revising hypotheses

4. Full Example: Abductive Analysis of Young v. Hitchens

4.1 Case Facts

Young v. Hitchens (fishing-net case)

Known facts:

  1. Plaintiff was fishing on the high seas with a net almost closed
  2. Defendant rowed into the net, caught the fish, and left

Legal rule (elements of theft):

  1. Performed a taking
  2. The property belonged to another
  3. Knew it belonged to another
  4. Had intent to take
  5. Exception: Taking to prevent danger

Phenomenon to explain: Did the defendant’s conduct constitute theft?


4.2 Hypothesis Generation (Applying Mill’s Five Methods)

Hypothesis 1: Lack of ownership knowledge (element 3 not established)

Method: Difference

Logical application:

Clearly marked net → Defendant knows ownership (usual case)
Unmarked net → Defendant may not know ownership (this case)
────────────────────────────────────────────
∴ Absence of marking causes absence of knowledge (sufficient-condition analysis)

Argument:

  • High seas are a public fishery; the net was not clearly marked
  • The net was “almost closed” but not fully closed; ownership unclear
  • Defendant may have sincerely believed the fish were a common resource

Supporting evidence:

  • Open-sea environment
  • No ownership marking
  • Net not fully closed

Potential weaknesses:

  • Nearly closed net suggests someone was working it
  • Defendant entered the net rather than going around

Defeasibility conditions: Evidence that defendant knew the net belonged to plaintiff (e.g., plaintiff’s loud warning)


Hypothesis 2: Possession incomplete (element 2 not established)

Method: Agreement

Logical application:

Case 1: Net fully closed + fish inside → Possession established
Case 2: Fish already caught on board → Possession established
This case: Net almost closed + fish still in water → ?
────────────────────────────────────
∴ “Complete control” is a necessary condition of possession (missing here)

Argument:

  • Fish still in the water, not fully removed from the natural state
  • Plaintiff’s control not yet complete
  • Under high-seas principles, incomplete possession does not exclude others’ intervention

Supporting evidence:

  • Net “almost” rather than “already” closed
  • Fish still in open water
  • Plaintiff had not yet actually extracted them

Potential weaknesses:

  • Plaintiff had already invested labor
  • Defendant’s conduct was predatory

Defeasibility conditions: Proof that industry custom treats “almost closed” as constituting possession


Hypothesis 3: Accidental conduct (element 1 or 4 not established)

Method: Residues

Logical application:

Defendant’s conduct = intentional theft + accidental entry
Known: No evidence of intent
────────────────────
∴ Remainder explained as “accident”

Argument:

  • Poor visibility; defendant did not see the gear clearly
  • Navigation error led to entry
  • Catching fish was opportunistic, not premeditated

Supporting evidence:

  • Weather / sea conditions
  • Defendant’s course
  • No evidence of premeditation

Potential weaknesses:

  • Defendant actively caught and took the fish away
  • No immediate apology or compensation

Defeasibility conditions: Evidence that defendant had previously observed the net’s location


Hypothesis 4: Necessity / emergency (exception applies)

Method: Concomitant variation

Logical application:

Low danger  → No need to enter the net
High danger → Must enter the net (to avoid peril)
────────────────────────────
∴ Degree of danger co-varies with necessity of entry

Argument:

  • Defendant’s vessel faced collision risk
  • Entering the net was the only way to avoid greater harm
  • Consistent with emergency-avoidance principles

Supporting evidence:

  • Sea conditions at the time
  • Positions of other vessels
  • Limits on defendant’s vessel control

Potential weaknesses:

  • Must prove “imminent danger”
  • Catching fish disproportionate to necessity of avoidance

Defeasibility conditions: Proof of no collision risk, or that defendant could safely go around


Hypothesis 5: Insufficient evidence (element 4 not established)

Method: Joint method

Logical application:

Agreement: All theft cases require “intent”
Difference: Intent → theft established; no intent → theft not established
────────────────────────────────────
∴ “Intent” is both necessary and sufficient
This case: Intent cannot be proven

Argument:

  • High-seas fishing is highly competitive; conduct may be resource-grabbing rather than theft
  • Cannot exclude the possibility of “mistaken belief of no owner”
  • Under criminal / strict standards, insufficient evidence should be read in favor of the defendant

Supporting evidence:

  • High-seas fishing background
  • No prior communication record
  • Conduct open to multiple readings

Potential weaknesses:

  • The conduct itself suggests intent
  • Defendant’s gain

Defeasibility conditions: Direct evidence of defendant’s intent (e.g., prior statements)


4.3 Hypothesis Evaluation Matrix

HypothesisExplanatory powerSimplicityTestabilityBackground fitOverall
H1 Knowledge gap⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐4.5
H2 Possession ambiguous⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐3.5
H3 Accidental conduct⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐3.0
H4 Necessity⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐2.5
H5 Insufficient evidence⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐⭐4.0

4.4 Inference to Best Explanation

Preferred hypothesis: H1 (lack of ownership knowledge)

Reasons:

  • Explanatory power: Fully explains why defendant entered the net (sincere belief in common resource)
  • Simplicity: No need to posit complex motives; based on objective environment
  • Testability: Verifiable through environmental evidence (markings, visibility)
  • Background fit: Fits the special high-seas fishing context

Confidence: Medium–high

Alternative hypothesis: H5 (insufficient evidence)

When applicable:

  • If defendant’s actual cognitive state cannot be proven
  • If a strict liability standard applies

Key uncertainties:

  • Defendant’s actual cognitive state (subjective intent)
  • Industry custom on the definition of “possession”

Verification suggestions:

  1. Investigate whether the net had clear markings
  2. Inquire into industry custom
  3. Investigate whether defendant received a warning at the time
  4. Analyze defendant’s past behavioral patterns

4.5 Hypothesis Revision Triggers

New evidenceHypothesis revision
Defendant observed the gear beforehandH3 overturned; shift to H1 or H5
Plaintiff loudly warned defendantH1 overturned; shift to H5
Industry custom: almost closed = possessionH2 strengthened
Defendant’s prior statements show intentH1, H3, and H5 all overturned
No collision risk at the timeH4 overturned

5. Another Example: Method of Difference in Tort

Case: Product Warning Label

Facts:

  • Case A: Warning label present → no harm
  • Case B: Warning label absent → harm occurs (this case)

Method of difference:

Warning present → no harm
Warning absent → harm
────────────────────
∴ Absence of warning is a sufficient condition of the harm

Abductive hypothesis:

  • Hypothesis: If a warning had been provided, the harm could have been avoided
  • Legal consequence: Manufacturer failed the duty to warn and bears tort liability

6. Output Template

markdown
# Abductive Reasoning Analysis Report

## 1. Phenomenon to Explain
[Describe the core dispute]

## 2. Known Facts
- [Fact 1] (source / strength)
- [Fact 2] (source / strength)

## 3. Competing Hypotheses (Mill’s Five Methods)

### Hypothesis 1: [description]
- **Method**: [agreement / difference / joint / residues / concomitant variation]
- **Logical application**: [concrete logical form]
- **Core claim**: [one sentence]
- **Support**: [evidence]
- **Weaknesses**: [counter-evidence]
- **Defeasibility conditions**: [overturning conditions]

### Hypothesis 2: [description]
...

## 4. Evaluation Matrix
[Scoring table]

## 5. Best Explanation
- **Preferred**: [hypothesis]
- **Reasons**: [based on criteria]
- **Confidence**: [high / medium / low]
- **Alternative**: [hypothesis]

## 6. Key Uncertainties
- [Gap 1]: [verification suggestion]

## 7. Revision Triggers
- [Evidence X] → [revision action]

7. Troubleshooting

Problem: Insufficient hypothesis generation

Solution: Mandatorily apply all five Mill methods; generate at least one hypothesis per method

Problem: Wrong method choice

Solution:

  • Find necessary elements → Method of Agreement
  • Find sufficient conditions → Method of Difference
  • Strengthen credibility → Joint Method
  • Explain remainder → Method of Residues
  • Degree relationships → Method of Concomitant Variation

Problem: Premature convergence

Solution: Must list at least 3 hypotheses and complete the evaluation matrix before choosing

Problem: Confusing reasoning types

Solution: Explicitly label “hypothesis” and “most probable explanation”; prepare revision triggers


8. Quick Checklist

  • Core phenomenon to explain has been identified
  • Known facts distinguished from information gaps
  • At least 3 Mill methods used to generate hypotheses
  • Each hypothesis has supporting evidence and counter-evidence
  • Structured evaluation matrix completed
  • Preferred and alternative hypotheses stated
  • Verification plan and revision triggers set
  • Defeasibility of hypotheses labeled

Frequently asked questions

What does the Legal Abductive Reasoning AI skill do?

Legal abductive reasoning skill for generating and evaluating the most reasonable explanatory hypotheses when evidence is incomplete or facts are ambiguous. Combines Mill’s five methods (agreement, difference, joint method, residues, concomitant variation) for structured causal inference. Triggered when the user needs to “propose possible legal explanations,” “generate defense/prosecution hypotheses,” “evaluate multiple readings of evidence,” or “infer the best explanation from incomplete facts.” Applies to case analysis, evidence evaluation, disputes over rule application, causal inference...

Why use Legal Abductive Reasoning on TypingMind?

Because you install it once and use it with any model. Legal Abductive Reasoning is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Legal Abductive Reasoning in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/THUYRan/Legal-Skills-Chinese/tree/main/English/skills/Legal-Abductive-Reasoning. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Legal Abductive Reasoning?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Legal Abductive Reasoning?

As many as you like. As long as a model supports skills, you can use Legal Abductive Reasoning with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Legal Abductive Reasoning AI skill free?

It is published on GitHub by THUYRan. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇