October workshops are openBuild a Search AI Agent$99 early bird

Enroll now
Skip to content

LAW & PUBLIC POLICY · CAPSTONE PROJECT

Contract comparison review agent

Build a source-linked comparison of two fictional service agreements. Verify candidate clause extractions and send missing or contradictory terms for review.

Beginner-friendly path · 10 guided lessons · editable JavaScript · runs in your browser · no API key

START WITH A QUESTION

Can a contract comparison catch its own mistaken extraction?

Predict how a value extracted into a table could differ from the original clause without anyone noticing.

Reveal what to look for

The workflow compares invented agreement clauses, rereads their sources, and blocks a briefing with a wrong extraction or mismatched version.

Start from the basicsExplore the path

10 guided lessons · learn at your pace

YOUR FINISHED BUILD

Contract comparison review agent

A clause comparison with exact source passages, verified differences, unresolved terms, and a reviewable local briefing.

WHY THIS FIELD USES THE IDEA

Contract review benefits from source-linked comparison tables that preserve missing terms and send interpretive decisions to a person.

SEE THE IDEAChoose a step

See the idea before you code it

Read the input

Two versioned fictional agreements, explicit termination-notice criteria, authored extraction mistakes, and reviewer approval.

function solve(input, tools) { /* your workflow */ }

Click a step to inspect this project's flow. The final workspace runs your JavaScript with original, authored records and checks.

A COMPLETE PATH TO YOUR BUILD

One small step at a time.

You can begin at the first lesson, resume shared skills you already practiced, or open the project IDE when you are ready.

01 · UNDERSTAND

Start with the essentials

No coding experience needed. These common lessons stay completed across every project.

  1. 1A model is one part of the systemFirst understand a model, a fixed workflow, and an agent that chooses its next tool call.
  2. 2Your first few lines of codeStart from zero: read values, objects, functions, and the result your code returns.
  3. 3Make a rule you can testWrite a small condition, change the input, and observe how the output changes.
  4. 4Turn an idea into a clear goalTurn the project's deliverable into a specific goal, allowed actions, and a visible stopping rule.

02 · PRACTICE

Learn the tools your project needs

Short explanations, clickable diagrams, and a small coding lab for each skill.

  1. 5A tool needs a clear contractKeep clause extraction and original-source verification as separate tools.
  2. 6Find a useful piece of informationRetrieve the exact agreement passage for each comparison row.
  3. 7Make a claim you can supportReport an omitted term as missing instead of filling it with a guess.
  4. 8Ask before changing somethingBind reviewer approval to a specific comparison draft.
  5. 9A proposal must earn acceptanceVerify candidate rule values before building a briefing preview.

03 · BUILD

Make your project work

Follow the project-specific ideas, insert the explained snippets, and inspect your real execution trace.

  1. 10Contract comparison review agentBuild and test your own version: A clause comparison with exact source passages, verified differences, unresolved terms, and a reviewable local briefing.

04 · TEST & REFLECT

Try to break it. Explain what holds.

3 runnable cases check the final project. Passing them demonstrates these examples, with limitations still to explore.

Change one thing

Change a candidate notice period while leaving its source passage unchanged. Verification must catch the mismatch.

Keep a useful learning record

Which questions require interpretation beyond the specific fields this comparison can verify?

Open project IDEReflect & save your notes
Inputs, checks, and project boundaries
Your starting material
Two versioned fictional agreements, explicit termination-notice criteria, authored extraction mistakes, and reviewer approval.
What your agent must check
  • Keep both exact source passages on an approved comparison.
  • Keep absent evidence distinct from contradictory extraction.
  • Do not mark a comparison complete with a mismatched source version.
Keep the scope clear
No real client documents, legal advice, enforceability judgments, representation, or actual publication. Explicit field checks do not perform legal interpretation.
Research behind this project path

Original explanations and authored practice records draw on these research and engineering ideas. The source organizations do not endorse this course or supply its fictional results.

Anthropic · Writing effective tools for agents — with agents

Design distinct tools with clear parameters, relevant returned information, and evaluations of how the agent actually uses them.

A description or schema does not guarantee the right action. A live tool can return different data for the same arguments as its environment changes.

Anthropic · Demystifying evals for AI agents

Define tasks, trials, and graders; inspect both execution records and final outcomes; repeat trials when model behavior varies.

A score depends on its cases and grading rules. Repeating a deterministic classroom case does not measure the variability of a live model.

OpenAI · Guardrails and human review

Distinguish automatic checks from approval decisions, pause sensitive tool requests, retain state, and resume after an application approves or rejects them.

Model-generated approval text is not authorization. Resume examples that automatically approve a request do not establish that a person reviewed it.

AWS · Amazon Bedrock AgentCore · Policy in Amazon Bedrock AgentCore: Control Agent Interactions

Evaluate identity and tool inputs at a gateway before allowing a call. Treat policy authoring, review, enforcement, and decision logging as separate operations.

The gateway governs capabilities routed through it. A classroom approval check is not an AgentCore integration or a complete production authorization system.

Your JavaScript really runs. The model decisions and school data are authored simulations, so you can learn without an API key. Every workspace also includes a separate real SDK example to explore next. Passing the lab’s cases is practice, not proof that an agent is ready for real-world use.