# SHEP: Legal Reasoning Platform for Law Students **URL:** https://shep.mintlify.app/ --- ## Table of Contents ### Get Started - [SHEP: Legal Reasoning Infrastructure for Law Students](https://shep.mintlify.app/introduction) - [Get started with SHEP: from sign-up to first submission](https://shep.mintlify.app/quickstart) - [Create and set up your SHEP account for the first time](https://shep.mintlify.app/account-setup) ### Core Concepts - [What are SHEP scenarios?](https://shep.mintlify.app/concepts/scenarios) - [Writing your legal analysis submission](https://shep.mintlify.app/concepts/submissions) - [SHEP chips: atomic units of legal reasoning](https://shep.mintlify.app/concepts/chips) - [Peer evaluation and grading in SHEP](https://shep.mintlify.app/concepts/peer-evaluation) ### Workflows - [How to analyze a legal scenario in SHEP](https://shep.mintlify.app/workflows/analyze-scenario) - [Grade and judge submissions in SHEP](https://shep.mintlify.app/workflows/grading) - [Practice legal reasoning with SHEP drills](https://shep.mintlify.app/workflows/practice) - [Using the SHEP Law Library for legal research](https://shep.mintlify.app/workflows/library) ### Your Account - [Manage your SHEP profile and public identity](https://shep.mintlify.app/account/profile) - [Configure notifications, privacy, and appearance](https://shep.mintlify.app/account/settings) - [Points, ledgers, and the SHEP leaderboard](https://shep.mintlify.app/account/points-leaderboard) ### Help - [Frequently asked questions about SHEP](https://shep.mintlify.app/help/faq) - [Troubleshooting common problems in SHEP](https://shep.mintlify.app/help/troubleshooting) ### External Links - [SHEP GitHub Repository](https://github.com/airobal/sskit-v1) - [SHEP Main Website](https://sheplegal.com) - [SHEP Support Email](mailto:students@sheplegal.com) --- ## Home Page SHEP is a legal-reasoning platform designed for law students who want to do more than memorize doctrine. You work through realistic legal scenarios, write structured analyses, annotate your reasoning into atomic units called chips, and evaluate your peers using professional rubrics. Over time, your contributions feed a growing knowledge graph that reveals where legal reasoning converges—and where it fractures. ### How SHEP works 1. **Register and set up your account** - Sign up at [sheplegal.com](https://sheplegal.com), complete onboarding, and configure your profile. 2. **Choose a scenario** - Browse scenarios across Foundational, Applied, and Frontier tiers. Each scenario presents a fact pattern for you to analyze. 3. **Write your submission** - Use the structured editor to write your legal analysis. Annotate key passages into chips—rules, cases, statutes, facts, and issues—that become part of the knowledge graph. 4. **Evaluate peers and earn points** - Grade other analysts' submissions using the rubric-based evaluation workflow. Earn XP and points that reflect your contribution to the platform. **Note:** SHEP prioritizes human reasoning. Every analysis, evaluation, and chip on the platform comes from a human mind—not an AI model. --- ## Introduction: SHEP: Legal Reasoning Infrastructure for Law Students SHEP is a legal-reasoning platform built for law students who want to practice the craft of analysis, not just memorize doctrine. Rather than generating answers for you, SHEP captures your reasoning — every issue you identify, every rule you apply, every argument you construct — and structures it as durable, queryable data. Every analysis on the platform comes from a human mind. No AI writes your submissions or grades your work. ### What you can do on SHEP **Analyze scenarios** - Work through realistic legal hypotheticals across Foundational, Applied, and Frontier tiers. Each scenario presents a fact pattern and prompts you to reason through the issues systematically. **Write submissions** - Draft structured legal analyses in the submission editor. Your writing is the core artifact — the raw reasoning that everything else builds on. **Build with chips** - Annotate your submissions into atomic reasoning units called chips: issue statements, rule statements, key facts, relevant cases, and argument links. Chips become part of the platform's knowledge graph. **Evaluate peers** - Grade other analysts' submissions using a rubric-based evaluation workflow. You assess methodology, not just conclusions — and earn points for doing it well. **Practice drills** - Run focused, timed exercises on specific doctrines or fact patterns. Drills let you sharpen reasoning on demand without the full scenario workflow. **Explore the knowledge graph** - Browse the structured corpus of rules, cases, facts, and arguments accumulated across all submissions. See where reasoning converges and where it fractures. ### Core principles SHEP is built around three ideas that set it apart from generic legal writing tools. **Human reasoning only** - Every submission, every evaluation, and every chip on the platform is produced by a person applying legal judgment. SHEP does not substitute AI-generated content for human analysis. **Provenance preservation** - SHEP tracks who asserted what, when, and on what basis. The chain of reasoning stays intact — not just the conclusion, but the structure of the argument and the authorities it relied on. **Structured disagreement** - When analysts reach different conclusions on the same scenario, SHEP treats that disagreement as first-class data. Conflicting analyses coexist and remain visible rather than being flattened into a single answer. ### Who SHEP is for SHEP is designed for law students who are actively working through legal analysis — in class, during bar prep, or as an independent practice habit. If you want structured feedback, a record of your reasoning over time, and a community of analysts working through the same novel legal questions, SHEP is built for you. SHEP stands for **Student Hypotheticals & Emergent Precedent**. The platform's long-term goal is a Virtual Precedent Repository where collective legal reasoning converges to reveal patterns in doctrinal evolution. ### Get started - [Quick start](https://shep.mintlify.app/quickstart) - Create your account and complete your first scenario analysis in minutes. - [Account setup](https://shep.mintlify.app/account-setup) - Learn how registration, email verification, and account approval work. --- ## Quick Start: Get started with SHEP: from sign-up to first submission SHEP gets you reasoning through real legal scenarios as quickly as possible. This guide covers everything from creating your account to submitting your first analysis — you can complete the entire flow in under fifteen minutes on your first visit. ### Account Approval New accounts go through an approval process before you can access the full platform. If you register with a .edu email address, your account is approved automatically. Non-.edu accounts require admin review — you will receive an email notification once your account is approved. ### 1. Create your account Go to [sheplegal.com](https://sheplegal.com) and click Start Reasoning to open the registration page. You can sign up in two ways: **Email and password** — Enter your full name, email address, and a password (minimum 8 characters, at least one letter and one number). Accept the Terms of Service and Privacy Policy, then click Sign up. **Google** — Click Sign up with Google, accept the terms in the consent modal, then complete the Google authentication flow. After you submit, SHEP sends a verification email to the address you provided. Keep that tab open while you check your inbox. ### 2. Verify your email Open the verification email from SHEP and click the confirmation link. The link brings you back to SHEP and confirms your address. Once your email is verified, you will see a confirmation message on the sign-in page. If the email does not arrive within a few minutes, check your spam folder. If you signed up with Google, email verification is handled by Google — you can skip directly to signing in. ### 3. Wait for account approval (non-.edu accounts only) If you registered with a non-.edu email address, your account enters a Pending state after verification. SHEP's team reviews new accounts and sends an approval email when your account is ready. Accounts registered with a .edu address are approved automatically and can sign in immediately after verification. ### 4. Complete onboarding When you sign in for the first time, SHEP walks you through a brief onboarding flow: - Set your display name and profile details - Choose your primary area of legal study (optional, helps SHEP surface relevant scenarios) - Review how submissions, chips, and peer evaluation work Complete each step and click through to your dashboard. You only see the onboarding flow once — you can update your profile at any time from Account > Profile. ### 5. Browse scenarios and write your first submission From your dashboard, navigate to Scenarios to see the scenario library. Scenarios are organized into three tiers: - **Foundational** — established doctrine applied to familiar fact patterns - **Applied** — developing areas where doctrine is still being shaped - **Frontier** — novel scenarios involving emerging technologies and unresolved legal questions Pick a scenario that interests you and open it. Read the fact pattern, then click Start Submission to open the submission editor. Write your analysis — identify the key issues, state the applicable rules, apply them to the facts, and reach a conclusion. When you are satisfied, click Submit. Your first submission does not need to be perfect. The platform is designed for deliberate practice — you will receive structured feedback through the peer evaluation system after submission. ### What happens after your first submission Once you submit, your analysis enters the peer evaluation queue. Other analysts on the platform will grade your submission using a rubric that assesses issue spotting, rule application, argument structure, and analytical depth. You earn points both for submitting and for evaluating others. As you annotate your submissions into chips — tagging rules, cases, facts, and issues — your reasoning becomes part of SHEP's knowledge graph, connecting your analysis to the broader corpus of human legal reasoning on the platform. **Related:** - [Understanding chips](https://shep.mintlify.app/concepts/chips) - Learn how to annotate your submissions into structured reasoning units. - [Peer evaluation](https://shep.mintlify.app/concepts/peer-evaluation) - See how the rubric-based grading workflow works and how to evaluate others. --- ## Account Setup: Create and set up your SHEP account for the first time Before you can analyze scenarios and submit your reasoning on SHEP, you need an approved account. This page explains how registration works, what to expect during the approval process, and how to complete your initial profile setup. The whole process takes a few minutes for .edu accounts and may take a short additional wait for non-.edu accounts while your account is reviewed. ### Register for an account SHEP offers two ways to create an account: email and password, or Google OAuth. Both options are available from the registration page at [sheplegal.com/register](https://sheplegal.com/register). #### 1. Open the registration page Go to [sheplegal.com](https://sheplegal.com) and click Start Reasoning to open the registration page. If you are already signed in, you will be redirected to your dashboard automatically. #### 2. Choose your sign-up method **Email and Password** Fill in the registration form with: - Full name — your display name on the platform - Email address — used for verification and notifications - Password — minimum 8 characters, must include at least one letter and one number Check the box to accept the Terms of Service and Privacy Policy, then click Sign up. **Google** Click Sign up with Google. A consent modal appears asking you to accept the Terms of Service and Privacy Policy before proceeding. Check the box and click Continue with Google to complete the OAuth flow with your Google account. If an account already exists for your email address, SHEP will show an error and direct you to sign in or reset your password instead. #### 3. Verify your email address After submitting the registration form, SHEP sends a verification email to the address you provided. Open the email and click the confirmation link to verify your address. Once verification is complete, you will see a success message on the sign-in page confirming that your email has been verified. Your account cannot progress to approval until your email address is verified. If the email does not arrive within a few minutes, check your spam or junk folder. Google sign-ups do not require a separate email verification step. #### 4. Wait for account approval SHEP uses a two-stage account status flow: Pending and Approved. - .edu email addresses are approved automatically at registration. You can sign in as soon as your email is verified. - Non-.edu email addresses enter a Pending state after verification and require admin review before you can access the platform. If your account is pending, you will receive an email from SHEP once a team member approves it. There is no action required on your part — just wait for the notification. If you signed in before approval, you will see a message indicating that your account is pending review. #### 5. Sign in Once your account is approved, go to [sheplegal.com/login](https://sheplegal.com/login) and sign in with your email and password, or click Sign in with Google if you used the Google sign-up flow. If you forgot your password, click Forgot password? on the sign-in page to request a reset link. ### Complete your profile during onboarding The first time you sign in with an approved account, SHEP guides you through a short onboarding flow to set up your profile. Onboarding covers your display name, optional study focus area, and a brief orientation to the platform's core concepts — scenarios, submissions, chips, and peer evaluation. You only see the onboarding flow once. After completing it, you land on your dashboard and can begin browsing scenarios. ### Manage your account settings After onboarding, you can update your profile and preferences at any time: - **Profile** - Update your display name, study focus, and public profile details. - **Settings** - Manage notification preferences, linked accounts, and other account options. ### Frequently asked questions **Why hasn't my account been approved yet?** Non-.edu accounts are reviewed manually by the SHEP team. Approval typically happens within one to two business days. If you have been waiting longer than that, contact support at [students@sheplegal.com](mailto:students@sheplegal.com). **Can I use a non-university email address?** Yes. SHEP accepts any valid email address at registration. Accounts registered with non-.edu addresses go through manual approval rather than being approved automatically. **I signed up with Google but I can't sign in. What should I do?** If your Google account is associated with a non-.edu address, your account may still be pending approval. Check your inbox for a notification from SHEP. If you see an error message at sign-in indicating a pending approval or account conflict, contact [students@sheplegal.com](mailto:students@sheplegal.com) for help. **How do I reset my password?** On the sign-in page at [sheplegal.com/login](https://sheplegal.com/login), click Forgot password? and enter your email address. SHEP will send a password reset link to your inbox. **Can I link both email/password and Google to the same account?** Account linking options are managed from your account settings. Contact [students@sheplegal.com](mailto:students@sheplegal.com) if you encounter a conflict between sign-in methods. --- ## Scenarios: What are SHEP scenarios? Scenarios are the core unit of work in SHEP. Each scenario presents a legal fact pattern — a set of parties, facts, and implied disputes — that you analyze by writing a submission. SHEP ships with a growing library of scenarios organized by legal domain and reasoning tier, so you always have a clear next challenge regardless of where you are in your legal education. Every piece of work you produce on SHEP — from a first submission to a peer evaluation — begins with a scenario. ### What a scenario contains Every scenario in the library includes: - **Title** — a concise name for the fact pattern (e.g., "The Palsgraf Problem," "Rideshare Driver Negligence") - **Facts** — the narrative of the dispute, written to surface relevant legal doctrine - **Issues** — pre-identified legal questions embedded in the fact pattern - **Domain** — the area of law the scenario belongs to (Torts, Contracts, Property, Criminal Law, Civil Procedure, Corporations, and more) - **Length bucket** — short, medium, or long, indicating the depth of analysis the fact pattern calls for Scenarios come from two sources: the Case Bank (the curated SHEP library) and Course Scenarios (fact patterns tied to your Paddock course syllabus). Use the Source filter on the Scenario Library page to switch between them. ### Reasoning tiers Every scenario in SHEP is classified into one of three reasoning tiers that represent a spectrum from classic, well-settled doctrine to genuinely novel, technology-driven disputes. The tier reflects how much established precedent can cleanly resolve the fact pattern — and therefore how much genuine legal uncertainty your analysis must confront. #### FOUNDATIONAL — Classic doctrine Foundational scenarios present canonical legal problems with settled doctrine. These are the kinds of fact patterns that have appeared on law school exams and in casebooks for generations. **Characteristics:** - Grounded in established doctrine (e.g., proximate cause in tort, offer and acceptance in contracts) - Always anchored to a specific doctrinal concept that ties the scenario to its area of law - No technology drivers — the legal question does not depend on any novel technology - Tests whether you can correctly identify and apply the controlling rule **Example:** The Palsgraf Problem — a proximate cause fact pattern grounded in classic torts doctrine, requiring you to reason through foreseeability and the scope of duty. Start here if you are new to SHEP or to a particular doctrine. Foundational work builds the analytical vocabulary — issue spotting, rule application, argument structure — that APPLIED and FRONTIER scenarios demand. #### APPLIED — Modern context Applied scenarios take established legal doctrine and place it inside a contemporary or industry-specific context. The underlying rules are still familiar, but the facts require you to reason about how courts adapt doctrine to modern circumstances. **Characteristics:** - Built on recognizable doctrine applied to a modern factual setting - Technology drivers are optional; when present, they must appear meaningfully in the facts — not just as background context - Doctrine anchor is optional but common - Requires you to transfer doctrine from its original context to new settings **Example:** Rideshare Driver Negligence — a gig-economy fact pattern that asks how respondeat superior and independent contractor doctrine apply to platform-based work relationships. Applied scenarios reward your ability to see how traditional rules fit non-traditional facts — a skill that matters in almost every practice area. #### FRONTIER — Futuristic and technology-driven Frontier scenarios present genuinely novel legal disputes where emerging technology creates questions that existing doctrine cannot resolve cleanly. No clear precedent applies. These are the scenarios SHEP most wants human analysts to reason through. **Characteristics:** - At least one futuristic technology driver is required (e.g., agentic AI, smart contracts/DeFi) and must be indispensable to the legal question — not just background color - Includes an indispensability explanation: why this fact pattern raises a question doctrine has never had to answer - Doctrine anchors are not required; the legal landscape is intentionally unsettled - Ordinary-dispute patterns in disguise are explicitly excluded from this tier **Example:** Autonomous Drone Delivery Dispute — a liability gap scenario where an AI-controlled drone causes harm and no existing principal-agent or products liability framework clearly resolves who is responsible. Frontier scenarios are where SHEP's purpose is clearest: capturing how trained analysts reason through problems the law has not yet solved. ### Progression model The three tiers reflect a learning path, not just a classification system. FOUNDATIONAL scenarios test whether you can apply established doctrine correctly. APPLIED scenarios test whether you can adapt that doctrine to unfamiliar facts. FRONTIER scenarios challenge you to reason in genuine uncertainty, where the answer cannot simply be looked up. As you complete submissions and earn evaluations across all three tiers, your profile reflects your analytical range and depth. Students progress from FOUNDATIONAL to FRONTIER as they mature on the platform — the tiers are designed to grow with you. ### Legal domains Scenarios are organized by legal domain. The Scenario Library groups them under doctrine headings so you can browse by area of law. The seven 1L curriculum domains are represented at the Foundational tier, with more advanced domains available at Applied and Frontier levels: - **Torts** - Negligence, strict liability, intentional torts, proximate cause, duty - **Contracts** - Formation, performance, breach, remedies, UCC Article 2 - **Property** - Possession, adverse possession, easements, covenants, future interests - **Criminal Law** - Elements, mens rea, defenses, inchoate offenses, sentencing - **Civil Procedure** - Jurisdiction, pleading, discovery, joinder, preclusion - **Corporations** - Formation, fiduciary duties, governance, piercing, liability Additional domains include Securities, Evidence, Constitutional Law, Labor Law, Agency, and Class Actions. ### Browsing and selecting scenarios Navigate to the Scenario Library from the main navigation. The library loads all available scenarios grouped by doctrine and supports filtering, search, and bookmarking. #### 1. Open the Scenario Library Navigate to Scenarios from the main navigation. The library loads all available scenarios grouped by doctrine heading. #### 2. Filter by doctrine, source, or length Use the filter bar to narrow by Doctrine (area of law), Source (Case Bank or Course Scenarios), and Length (short, medium, or long). Filters combine — you can select Torts + Short to find compact doctrine exercises, for example. #### 3. Search by keyword Use the search bar to find scenarios by title, facts text, or domain name. Results update as you type without reloading the page. #### 4. Expand a doctrine group Click any doctrine heading to expand the scenarios within it. Each card shows the scenario title, a preview of the facts, and up to three issue tags. If a scenario has more issues, the card shows how many remain. #### 5. Bookmark scenarios for later Hover over any scenario card and click the bookmark icon to save it. Your bookmarked scenarios persist across sessions and appear with a filled bookmark icon when you return. #### 6. Open a scenario and begin your submission Click a scenario title to open its detail page, read the full fact pattern, and begin your written analysis. See [Submissions](https://shep.mintlify.app/concepts/submissions) for how the writing workflow works. Course scenarios are assigned by your instructor through Paddock, the course management layer of SHEP. If your institution uses Paddock, switch the Source filter to Course Scenarios to see the fact patterns relevant to your current module. --- ## Submissions: Writing your legal analysis submission A submission is your written response to a scenario. It is the primary artifact you produce in SHEP: a structured legal analysis where you identify the issues, apply relevant rules and cases, construct arguments, and reach a conclusion. Every submission is tied to one scenario, belongs to you as its author, and moves through a lifecycle from draft to submitted to evaluated. Submissions are not disposable practice. SHEP treats each one as a first-class, durable piece of human legal reasoning. Who wrote it, when, and what arguments it made — that provenance is preserved in the platform's knowledge infrastructure and contributes to convergence analysis over time. ### The structured editor When you open a scenario and begin writing, you work inside SHEP's rich-text editor. The editor supports standard document formatting — headings, paragraphs, lists — and includes annotation support that lets you highlight specific passages to create chips. Chips are atomic reasoning units extracted from your text: a rule you cited, a case you applied, an issue you identified, a key fact from the scenario. When you annotate a passage, you are not just marking it up for display — you are converting that text into a structured data point that SHEP can aggregate, compare across submissions, and surface in the knowledge graph. You do not need to think about chips while you write. Annotation works alongside or after drafting. Focus on the analysis first; the structure follows from your highlights. ### Drafts vs. submission You can save your work at any point as a draft. Drafts are private — they are not visible to other analysts and are not available for peer evaluation until you formally submit. There is no time limit on drafts; you can return across multiple sessions and continue editing before you commit. #### 1. Write your analysis Open a scenario and write your legal analysis in the editor. Use headings and paragraphs to organize your issue spotting, rule application, and argument construction. Write as you would for a legal memo or exam answer. #### 2. Annotate key passages Highlight important text — rule statements, case citations, issue formulations, key facts — to create chips. Each highlight becomes a structured reasoning unit attached to your submission. You can annotate as you draft or return to annotate after the analysis is complete. #### 3. Save as draft Use the save action to preserve your work in progress. The editor autosaves periodically, but you can also save manually at any time. Drafts remain private until you submit. #### 4. Submit When your analysis is ready, submit it. Submitting makes your work available to other analysts for peer evaluation. You become eligible to receive rubric-based scores, per-criterion feedback, and narrative comments from evaluators. Once you submit, your analysis enters the peer evaluation queue and becomes visible to evaluators. You can continue to revise a submitted analysis, but evaluations already completed against earlier versions are preserved in the record. ### One submission per scenario You have one submission per scenario. You can revise that submission as many times as you want while it is in draft. After submission, you can still update and refine your analysis, but SHEP tracks revisions so the system and evaluators can distinguish the version that was graded from later edits. This design is intentional. It reflects how legal analysis works in practice: you apply your best judgment, commit to a position, and receive feedback — rather than spinning up unlimited parallel drafts that dilute accountability. ### How your submission feeds into peer evaluation When you submit an analysis, it enters the peer evaluation queue. Other qualified analysts on SHEP can select your submission and evaluate it: scoring it against a structured rubric, validating the accuracy of your chips, and writing narrative feedback that identifies what your analysis did well and what it should improve. The number of evaluations your submission receives, and how much those evaluations converge in their scores and chip assessments, builds a confidence signal in the SHEP knowledge graph. Submissions that many evaluators assess similarly — in their rubric scores and in how they characterize the reasoning — carry a stronger convergence signal. Submissions where evaluators disagree significantly preserve that disagreement as structured data. **Related:** - [Chips](https://shep.mintlify.app/concepts/chips) - Learn how your annotations become structured reasoning units that enter the knowledge graph. - [Peer evaluation](https://shep.mintlify.app/concepts/peer-evaluation) - Understand how other analysts review and score your submission. ### Viewing your submission After submitting, your submission's detail page shows: - Your full analysis text, rendered with your formatting - Your annotated issue statements displayed as ISSUE chips below the main content - Your case citations, with verification status for each one (verified, unverified, not found, or ambiguous) - Any evaluations you have received, including per-criterion scores, overall score out of 100, and written feedback from each evaluator - Reaction counts from other analysts who found your analysis useful - A knowledge graph view that visualizes the reasoning relationships between your chips as a directed argument flow diagram The detail page is also where you manage citation verification for your case and statute highlights. You can verify citations individually using the look-up tool, batch-verify all unverified citations at once, or resolve ambiguous matches against the citation database. After submitting, go to your submission's detail page and verify your case citations. Verified citations carry more weight in convergence analysis and signal careful, accurate legal research to evaluators. ### Submission types SHEP supports more than one type of analysis for a single scenario. In addition to a standard analytical memo, you may have the option to produce a litigation-style analysis for the same fact pattern — approaching the scenario from an advocacy framing rather than a neutral analytical one. Completing multiple submission types for the same scenario earns additional points and deepens your analytical record on that fact pattern. Your submission's detail page shows which submission types you have completed and which remain available for the current scenario. A What's Next? prompt appears when additional work types are open for you to complete. --- ## Chips: SHEP chips: atomic units of legal reasoning A chip is the smallest meaningful unit of legal reasoning in SHEP. When you write a submission and highlight a passage — a rule you cited, a case you applied, a factual assertion, an issue you identified — that highlight becomes a chip. Chips are the mechanism by which your written prose is converted into structured, queryable data that persists beyond any individual submission and becomes part of SHEP's growing legal knowledge graph. Chips capture not just what you wrote, but what kind of reasoning it represents. A chip knows whether it is a rule statement, a case citation, a statutory provision, a key fact, or an issue formulation. This classification is what lets SHEP aggregate chips across many submissions, identify where analysts converge, and build network relationships between legal concepts. ### Chip types Chips fall into three groups based on how they are accepted and what they represent. #### Authority chips Authority chips reference external legal sources. Because they make claims about the real world — about what a statute says, what a case held, what a regulation requires — SHEP subjects them to a verification process. An authority chip starts as unverified and becomes verified when its citation is confirmed against the legal authority database. | Type | What it represents | |------|-------------------| | RULE | A statement of a legal rule drawn from a primary or secondary source | | CASE | A citation to a decided case | | STATUTE | A citation to a legislative enactment | | REGULATION | A citation to a regulatory provision | | GUIDANCE | Agency guidance, interpretive rules, or informal policy statements | | TREATISE | A citation to a legal treatise | | ARTICLE | A citation to a law review article or academic publication | | POLICY | A reference to a policy document, corporate rule, or institutional policy | #### User chips User chips represent your own analytical work — things you identified or asserted in your analysis. These do not reference external sources and are accepted automatically without a verification step. | Type | What it represents | |------|-------------------| | FACT | A key factual assertion from the scenario that you are treating as legally significant | | ISSUE | A legal issue you have identified and framed | #### Additional chip types These types capture more specific aspects of legal reasoning and argument structure. | Type | What it represents | |------|-------------------| | HOLDING | The specific holding of a cited case | | ELEMENT | A required element of a legal test or cause of action | | FACTOR | A discretionary factor in a balancing test or multi-part analysis | | PRINCIPLE | A broad legal principle that informs but does not directly resolve the issue | | SECONDARY_SOURCE | A secondary source other than a treatise or article (e.g., a restatement section) | | EVIDENCE | Concrete supporting material such as exhibits or records, treated as a fact-like chip | EVIDENCE chips are user-authored reasoning — they do not require verification. They are distinct from CASE or STATUTE chips, which assert claims about external legal sources. ### Verified vs. unverified chips For authority chips, verification status matters. When you cite a case or statute in your submission and create a chip for it, that chip starts in an unverified state. SHEP can attempt to verify the citation automatically against CourtListener and other integrated databases. You can also look up a citation manually using the citation search panel on your submission's detail page. A verified chip means the citation has been confirmed as a real legal source that exists in the database. A not-found chip means the citation could not be matched — which may indicate a citation error in your analysis. An unverified chip has not yet been checked. Verification status is visible to evaluators and affects how your chips contribute to the knowledge graph. Unverified authority chips are captured as reasoning, but they carry less weight in convergence analysis until confirmed. After you submit, check the case citations section of your submission's detail page. You can batch-verify all your case citations at once, or look up individual ones to resolve ambiguous matches. ### Relations between chips Chips do not exist in isolation. SHEP supports explicit relationships between chips that express how pieces of legal reasoning connect to each other. These relations are what make the knowledge graph a graph rather than just a list of annotations. | Relation | Meaning | |----------|---------| | SUPPORTS | One chip provides support for another | | CONTRADICTS | One chip is in tension with or contradicts another | | APPLIES_TO | A rule or principle chip applies to a fact or issue chip | | DISTINGUISHES_FROM | One case or rule chip is distinguished from another | | REBUTS | One chip rebuts a claim made in another | | DEPENDS_ON | One chip's validity depends on another chip being accepted | | EVIDENCE_FOR | A fact or evidence chip provides support for a rule or holding chip | | SIMILAR_TO | Two chips express analogous reasoning in different contexts | Relations are created through the annotation and argument-linking workflow inside the submission editor. When you mark an argument span and link chips to it, SHEP captures the relational structure of your reasoning — not just the isolated assertions. ### How chips contribute to the knowledge graph Every chip you create — verified or not — enters SHEP's legal knowledge graph as a node. When multiple analysts submit analyses of the same or related scenarios, their chips can be compared, clustered, and evaluated for convergence. If many analysts cite the same rule and apply it to the same kind of fact pattern, that convergence signal strengthens the chip's authority status in the graph. If analysts disagree — citing different rules, reaching different conclusions — that disagreement is preserved as structured data rather than collapsed into a single answer. This is the core purpose of the chip system: to build a machine-readable, human-authored record of how legal reasoning actually gets applied, including where it is stable and where it fractures. ### Managing your chips Your personal chip library is accessible from the My Chips page. This page shows all chips you have saved across all submissions, organized by type or folder. You can: - Search chips by content or type - Organize chips into named folders - Add personal notes to individual chips - Copy a chip with its citation formatted for external use - Insert a chip directly into an active draft from the research context bar **Related:** - [Submissions](https://shep.mintlify.app/concepts/submissions) - Learn how chips are created from your submission text. - [Peer evaluation](https://shep.mintlify.app/concepts/peer-evaluation) - See how evaluators assess chip accuracy and applicability during grading. --- ## Peer Evaluation: Peer evaluation and grading in SHEP Peer evaluation is the process by which submitted analyses get assessed on SHEP. When you submit your legal analysis, other analysts can read it and evaluate it — scoring it against a structured rubric, providing written feedback on what you did well and what needs improvement, and assessing the accuracy of the chips in your submission. In return, you can evaluate other analysts' submissions and earn points and XP for the quality of your evaluations. This system draws on a core insight: legal reasoning is best assessed by practitioners and students who are doing the same work. Peer evaluation in SHEP is structured enough to produce consistent, comparable scores, while still leaving room for substantive written feedback that reflects real analytical judgment. ### Two evaluation modes SHEP offers two distinct modes for engaging with another analyst's submission. These modes are fundamentally different and should not be confused. **Grade** - Score a submission against the rubric criteria, provide per-criterion notes and evidence links, and write narrative feedback identifying strengths and areas for improvement. Grading creates a review record against the submission. **Judge** - Write a judicial opinion or ruling on the underlying scenario — not an evaluation of the submission, but an independent legal determination of how the scenario should be decided. Judging creates a new submission of type judge, authored by you. Grade and Judge are distinct. When you grade, you are assessing another analyst's reasoning. When you judge, you are writing your own judicial opinion on the scenario itself. A judge submission is a first-class piece of authored work — not a review of someone else's. ### The grading rubric Every submission is evaluated against a structured rubric. Each rubric consists of criteria — named dimensions of legal reasoning quality — each with a weight that reflects its relative importance to the overall score. Within the grading workspace, you score each criterion on a 0–4 scale: | Score | Meaning | |-------|---------| | 0 | Absent or fundamentally flawed | | 1 | Attempted but significantly incomplete or incorrect | | 2 | Adequate — meets the basic standard | | 3 | Strong — above baseline with clear analytical depth | | 4 | Excellent — thorough, precise, and well-supported | Each criterion has an anchor description that tells you what a score of 0, 2, and 4 looks like for that dimension. You can expand any criterion in the rubric panel to read its description and the score anchors before selecting a score. Certain criteria are chip-validated: their scoring can be supported by linking specific highlights from the submission as evidence. When you score a chip-validated criterion, you can attach relevant chips to that criterion to justify your score and build rule links in the knowledge graph. Your per-criterion scores are combined with their weights to produce an overall score out of 100. ### The grading workflow #### 1. Select a submission to grade Navigate to the Grade section. SHEP surfaces submissions from analysts working in similar domains that need evaluation. You can see scarcity signals — domains with few active evaluators are highlighted so you can contribute where your review matters most. #### 2. Read the submission The grading workspace opens a full-screen modal with the submission on the left and the rubric on the right. Read the submission carefully before scoring. Switch to the Chips tab to see the structured chips the analyst annotated. #### 3. Score each criterion Work through the rubric, selecting a score (0–4) for each criterion. Expand a criterion to read its description and anchors. Add notes to explain your score for any criterion where context is helpful. #### 4. Validate chips In the Chips panel, assess each chip in the submission for accuracy and applicability. For each chip, you indicate whether it is accurate or has an issue, whether it applies to the scenario, and whether it is the right tier of authority (controlling, persuasive, non-legal, or wrong jurisdiction). You can also flag completeness issues — missing elements, missing exceptions, or missing standards — that the analyst should have included. #### 5. Write narrative feedback In the rubric panel, write bullet-point strengths and areas for improvement. This narrative is delivered to the submission author as qualitative feedback alongside the numeric scores. You can also add private notes that only you can see. #### 6. Submit the review Submit when your scoring and feedback are complete. SHEP checks whether you have completed chip validation and linked evidence for required criteria before allowing submission, and will warn you if significant items are missing. Once submitted, the author sees your scores and feedback. The grading workspace autosaves your progress every 90 seconds while you have the tab open. If you need to return to a review later, it is saved as a draft review that you can resume from the Grade page. ### Earning points and XP Completing a peer evaluation earns you points and XP. The amount depends on several factors: the quality and thoroughness of your review, the domain scarcity (evaluations in domains with few active reviewers earn a multiplier), and the completeness of your chip validation. Your points accumulate in the SHEP ledger and contribute to your standing on the leaderboards. The system is designed so that evaluating other analysts' work is as valued as producing your own. Careful, thorough evaluation is how you build analytical credibility on the platform. ### Convergence As multiple analysts evaluate the same submission, SHEP builds a convergence signal. When many evaluators independently reach similar rubric scores and characterize the reasoning in similar ways, the submission's analysis gains confidence in the knowledge graph. When evaluators disagree substantially, that disagreement is preserved as structured data. Convergence does not mean consensus. SHEP does not flatten disagreement into a single "correct" answer. Instead, it tracks where reasoning converges and where it diverges — preserving the structure of analytical agreement and dispute across the full population of evaluators. This is what allows SHEP to reveal, over time, where legal doctrine is stable and where it remains genuinely contested. **Related:** - [Scenarios](https://shep.mintlify.app/concepts/scenarios) - Understand the fact patterns that submissions respond to. - [Chips](https://shep.mintlify.app/concepts/chips) - See how chip accuracy assessment connects to the grading workflow. --- ## Workflows: How to analyze a legal scenario in SHEP Analyzing a scenario is the core workflow in SHEP. You read a fact pattern, write a structured legal analysis in the editor, annotate key reasoning units as chips, connect those chips with relational links, and submit for peer evaluation. This page walks you through every step. ### 1. Browse and pick a scenario Navigate to Scenarios or Library from the main navigation. The page groups scenarios by doctrinal area — Torts, Contracts, Property, Criminal Law, Civil Procedure, and more. Use the filters to narrow your options: - **Doctrine** — filter to a specific area of law (e.g., Torts, Contracts). - **Source** — choose Case Bank for platform scenarios or Course Scenarios for content tied to your enrolled course. - **Length** — Short, Medium, or Long fact patterns. Scenarios carry a reasoning tier that signals the type of legal challenge: | Tier | What it tests | |------|--------------| | FOUNDATIONAL | Classic doctrine — proximate cause, offer and acceptance, adverse possession | | APPLIED | Modern context — gig-economy relationships, platform liability, data privacy | | FRONTIER | Novel questions driven by emerging technology — AI liability, smart contracts, autonomous systems |