How to use the Customize Corporate Legal Profile agent skill

What is an agent skill, and what does customize do?

An agent skill is a set of instructions, packaged as a folder with a SKILL.md file, that an AI agent loads when a request matches it. The customize skill is one of these. Its job is to edit one setting in an existing corporate legal profile.

The agent reads your existing configuration files, shows you a summary of what is currently set, asks what you want to change, explains which other tools will be affected, and then writes the new value. It also warns you when a change makes the profile inconsistent or weakens built-in safety flags.

• Customizable areas include company details, active modules (M&A, Board & Secretary, Public Company, Entity Management), risk posture, people and escalation chain, materiality thresholds, output formats, and matter workspace paths.
• It changes one item at a time and does not re-ask the whole interview.
• It edits configuration only. It does not check the substance of legal positions.

Who should use it, and who should not

The skill is built for legal teams that already completed the corporate-legal setup interview and want to tweak a single setting. It also suits people who maintain a shared company profile used across several legal plugins.

If you have not run the cold-start setup yet, the skill stops and asks you to run setup first. If you expect it to give legal advice, it will not: it only edits configuration. Changes such as materiality thresholds can also alter what other legal tools flag.

What you need before installing

The skill requires the claude-for-legal corporate-legal plugin with a completed cold-start setup. Your config files should be in ~/.claude/plugins/config/claude-for-legal/. The skill reads corporate-legal/CLAUDE.md and company-profile.md in that folder.

According to the source, the skill does not run scripts and does not use the network. The package contains a single file, SKILL.md.

Install the customize skill

You can install the skill in either of two ways. Download the ZIP from the item page and unzip it into one of these locations:

• Personal skills folder: ~/.claude/skills/
• Project skills folder: .claude/skills/ inside your project
• Or, in the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.

Trigger the skill with natural requests

You can start the skill with the slash command /corporate-legal:customize, optionally followed by a section name or a description of what you want to change. The skill also responds to requests such as "change my [thing]", "update my profile", "edit my config", or "customize".

When it starts, it lists your profile grouped by area with the current value of each item, then asks what you want to adjust. Pick a section or describe the change in your own words.

• "Change my materiality threshold for contracts."
• "Update my profile: turn on the Public Company module."
• "Edit my config so my escalation contact is the GC."

Examples of changes and their downstream effects

Before writing any change, the skill shows the current value, asks for the new one, and explains what changes downstream. Examples from the skill's instructions:

• Materiality threshold from $250K to $500K: the diligence and material-contract skills treat $500K as the cutoff. Existing findings stay as logged; re-run if you want the new threshold applied retroactively.
• Turning on the Public Company module: the skill prompts for reporting calendar and disclosure controls the next time you run anything in that area.
• AI bulk-review trust from "check every row" to "spot-check 10%": the AI tool handoff skill will QA a 10% sample instead of every extraction.

Safety notes and guardrails

The skill never deletes a section. To remove something, it sets the item to [Not configured] and explains what that means for the plugin's behavior.

It flags inconsistencies, such as turning off the Public Company module while keeping SEC counsel in the escalation chain, or pairing an aggressive risk posture with a very low materiality threshold.

It also warns before you remove guardrails. The [review] flag, source attribution tags on retrieved documents, and [verify] tags on cited authorities are described as load-bearing, so the skill explains the trade-off before you remove them.

Changes to shared fields such as company name, industry, jurisdictions, practice setting or stage are written to company-profile.md, and the skill notes that all 12 plugins read that file.