How to use the Customize Commercial Legal Profile agent skill

What is an agent skill and what does this one do?

An agent skill is a folder containing a SKILL.md file that gives an AI agent specific instructions for a task. This skill covers one job: editing a saved profile for a commercial contracts review assistant.

The profile holds settings such as your risk tolerance, escalation contacts, standard contract positions, NDA triage rules, and output style. You name the thing you want to change, the agent shows the current value, asks for the new one, explains what will change in later outputs, and then saves it. It also warns you if a change conflicts with other settings or would weaken a safety check.

• Customizable areas include company details, risk posture, people and approvers, playbook positions, NDA triage, review preferences, house style, workflow paths, and integrations.
• Changes to company details are written to a shared company profile used by all 12 plugins in the collection.
• The skill never deletes a section. Removing one sets it to [Not configured].

Install the customize skill

Download the ZIP file for this skill from this page. Then choose one of the two install options below.

For a personal install available in all your projects, unzip the file into ~/.claude/skills/. For a single project, unzip it into the .claude/skills/ folder inside that project.

If you use the Claude apps, you can upload the ZIP instead. Go to Settings, then Capabilities, then Skills, and upload it there.

• Personal install: ~/.claude/skills/
• Project install: .claude/skills/ inside your project folder
• Claude apps: Settings → Capabilities → Skills

Check the requirements before you start

The skill needs Claude Code or a compatible agent with the claude-for-legal commercial-legal plugin installed. It also needs a completed cold-start setup that produces the plugin config files in ~/.claude/plugins/config/claude-for-legal/.

If setup has not been completed, or the config still contains [PLACEHOLDER] values, the skill stops and tells you to run /commercial-legal:cold-start-interview first. This skill adjusts an existing profile; it does not create one.

How to trigger the skill with natural requests

You can invoke the skill with the slash command /commercial-legal:customize, or describe what you want in plain words. The skill's description responds to requests such as "change my risk posture", "update my profile", "edit my playbook", "tune my config", or "customize".

The skill can also take a section name or a description as its argument. If you are unsure what is available, ask it to show the customizable map first. It will list each area with a one-line summary of its current value.

• "Change my risk posture to aggressive."
• "Edit my playbook: set the liability cap fallback to 6 months."
• "Update my profile: add a new escalation approver for high-value deals."
• "Tune my NDA triage preferences."
• "Customize my renewal alert channel."

What happens during a change

The skill reads your current configuration, shows the customizable map, and asks what you want to adjust. It then shows the current value, asks for the new one, and explains what changes downstream. You confirm before anything is written to the config.

Each session handles one change at a time, so the skill does not re-ask the whole interview. Examples of downstream effects described in the source include a new liability cap fallback being flagged as a deviation by the review command, and a new escalation approver being included by default for a matching risk band.

• Example: a liability cap fallback changed from 12 months to 6 months means anything above 6 months is flagged as a deviation. Existing deal-debrief entries stay as logged.
• Example: moving risk posture from middle to aggressive means vendor-friendly positions are accepted without flagging, and the review bar moves higher.
• Example: setting a value to [Not configured] tells you what that means for the plugin's behavior.

Safety notes: guardrails, scripts and network

The skill protects several safeguards. The [review] flag, source attribution tags, and [verify] tags on cited statutes are described as load-bearing and should not be removed. If you ask to turn one off, such as dropping the [review] flag, skipping the privilege header, or removing [verify] tags, the skill explains what the guardrail protects against and asks you to confirm the trade-off.

The skill does not run scripts and does not use the network, according to its listing. It reads and writes local configuration files only.

Changes alter how contracts are flagged. Outputs still need review by a qualified lawyer, and this skill is not a substitute for legal advice.

• Runs scripts: no
• Uses network: no
• It flags conflicts, such as an aggressive risk posture combined with a rule that every redline needs GC approval, and asks which you want.

Tips for getting reliable results

Change one thing at a time and check the summary the skill gives you before confirming. Review the downstream effects it describes, since they determine how future contract outputs will look.

If you share a profile across several plugins, edit company-level fields deliberately, because they affect every plugin that reads them. Keep a record of changes your legal team agrees to, especially for playbook positions.

• Name the section you want (for example, "NDA triage") to skip the menu.
• Describe the business reason for the change so the skill can check for conflicts.
• Re-run the customize skill any time you want to adjust the profile again.