How to use the Client Intake Assistant agent skill

What the client-intake skill makes the agent do

An agent skill is a set of instructions, stored in a SKILL.md file, that tells Claude how to handle a specific task. When this skill is installed, Claude follows a set workflow for a new client intake: it routes the client's problem to a practice area (immigration, housing, family or consumer), asks that area's intake questions, and listens for issues outside that area.

The skill then checks for conflicts of interest, suggests a triage level (Urgent, Time-sensitive, Standard or May be out of scope), and produces a formatted case summary with an AI-assisted label, verification prompts and supervision routing. Where an intake surfaces a deadline, it also emits copy-paste-ready /legal-clinic:deadlines --add blocks.

The skill is explicit that it does not decide whether the clinic accepts a case. That decision stays with the student's analysis and the professor's judgment.

• Routes the client's problem to a practice-area intake template
• Flags cross-area issues, such as immigration status raised in a housing intake
• Flags possible conflicts for professor review without resolving them
• Suggests a triage level as an input, not a acceptance decision
• Writes a case summary for a student to analyze and a professor to review

What you need before installing

The skill needs Claude Code or a compatible SKILL.md agent runtime. It also relies on a configured CLAUDE.md file at ~/.claude/plugins/config/claude-for-legal/legal-clinic/ that sets out your practice areas, intake templates and supervision settings. Without that file, the skill falls back to generic defaults.

Practice-area guides in the same config folder are optional. The workflow also references the companion /legal-clinic:deadlines and /legal-clinic:build-guide skills.

• Claude Code or a compatible SKILL.md runtime
• A configured CLAUDE.md at ~/.claude/plugins/config/claude-for-legal/legal-clinic/
• Optional: practice-area guides under guides/ in the same folder
• Optional companion skills: /legal-clinic:deadlines and /legal-clinic:build-guide

Install the client-intake skill

Download the ZIP from this page and unzip it. For a personal install, place the client-intake folder in ~/.claude/skills/. To scope it to one project, place it in that project's .claude/skills/ folder. The ZIP contains SKILL.md and references/intake-templates/README.md.

If you use the Claude apps instead, you can upload the skill under Settings → Capabilities → Skills.

• Personal install: ~/.claude/skills/
• Project install: the project's .claude/skills/ folder
• Claude apps: Settings → Capabilities → Skills

Set up your CLAUDE.md and intake templates

Before the first intake, create the configuration at ~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md. The skill loads it for practice areas, intake templates, supervision style, jurisdiction and flag triggers.

The references/intake-templates/README.md file explains that one template file is expected per practice area: immigration.md, housing.md, family.md and consumer.md. If your clinic has a practice-area guide, place it under guides/<practice-area>.md in the same config folder. The skill checks for it and uses its questions in place of the generic defaults.

• Config file: ~/.claude/plugins/config/claude-for-legal/legal-clinic/CLAUDE.md
• Template files: immigration.md, housing.md, family.md, consumer.md
• Optional guide: guides/<practice-area>.md

Trigger the skill with natural requests

The skill's description says to use it when starting a new client intake, running an intake interview, or writing up a new client's situation. You can invoke it by name with /client-intake, and the skill accepts an optional practice area hint as an argument. You can also make a plain request in the same spirit, such as asking Claude to run an intake for a new client.

The skill's first question to the client is open-ended: what brought them to the clinic today. The client does not need to know the legal category of their problem.

• Explicit: /client-intake
• With a hint: /client-intake [optional: practice area hint]
• Natural: "Start a new intake for a client who is worried about a landlord lockout."

Examples of the workflow in action

Suppose a housing client mentions their immigration status. The skill notes the housing issue, then flags an immigration issue as well, because the table in the skill treats that statement as an immigration issue even within a housing intake. A landlord who threatens to call ICE is listed as a housing, immigration and possible retaliation matter.

Suppose a consumer client describes a partner's threatening behavior. The skill flags a possible DV, family law or protective order issue alongside the consumer matter. In every case the flags go into the summary for the student and professor, and the clinic decides what to do with them.

• Housing intake where the client mentions immigration status: housing and immigration flags
• Consumer intake where the client describes threats from a partner: consumer, DV and family flags
• Intake where the client says a paycheck is being garnished: consumer and employment overlap flag

Safety notes: scripts, network and sensitive data

This skill runs no scripts and uses no network access, according to its listing. Its outputs are text: intake questions, the case summary, conflict flags and triage suggestions.

Intake involves immigration, safety and criminal history details. Before entering any client information into an AI tool, check your clinic's confidentiality and privacy rules. The skill's own template guidance notes that criminal history questions are sensitive and should come with an explanation of why they are asked.

• Runs scripts: no
• Uses network: no
• Check clinic confidentiality and privacy rules before entering client details
• The skill does not give legal advice and is built for supervised staff