# Personal Preferences Personal preferences are role-based instructions stored in each employee's Claude Enterprise profile. They tell Claude who the employee is, what kind of work they do, what help they need, and how to communicate with them effectively. Every conversation the employee has with Claude is shaped by these preferences automatically. *** ## How They Work Each employee's profile contains a structured prompt written in first person ("I am the...") and formatted in XML. When the employee opens any Claude conversation, Claude reads these preferences and adjusts its behavior accordingly. The preferences persist across all conversations, projects, and sessions. The prompt is organized into five XML sections that cover identity, role, cognitive style, operating instructions, and friction rules. Together they give Claude everything it needs to calibrate its behavior for that specific employee. The result: Claude already knows the employee's role, priorities, and communication style before they type a single message. *** ## Prompt Structure Each profile prompt is built from five XML sections: | Section | What It Contains | | ------------------------ | -------------------------------------------------------------------------------------------------------------------------------- | | `` | Who the employee is as a human. Identity signal translated into behavioral context. | | `` | Their role and day-to-day work, written as context Claude needs to situate its help. | | `` | Cognitive style. How they process hard problems, how they want feedback delivered, how they relate to being wrong. | | `` | Direct operating instructions: preferred response depth, tone, formality, swearing policy, and any specific behavioral requests. | | `` | Friction rules. What wastes their time or makes them shut down, and any personal context translated into behavioral guardrails. | Prompts are generated entirely from a conversational onboarding. There is no static role library required. Each employee's profile is built from what they share about themselves during the Personalizer conversation (see below). *** ## Personalizer **Status:** In Development Employees build and maintain their profile prompts through a conversational skill called Personalizer. Instead of filling out a form or selecting a role from a dropdown, employees trigger the skill with `/personalizer` in any Claude conversation and go through a short, warm, 10-to-15-minute conversation with Claude. ### How to Access The skill is available directly inside Claude. Employees type `/personalizer` in a Claude conversation to trigger it. No external link, no separate tool, no sign-in step beyond being in Claude Enterprise already. ### The Conversation The conversation covers 13 questions across five zones: ground truth (name, role, day-to-day), who you are (what lights you up, what makes work meaningful, when you are most locked in), how you think (problem-solving style, feedback preferences, how you handle being wrong), what you need from Claude (depth, tone, formality, swearing), and friction rules (what wastes your time, any personal context that affects your work). Claude reacts between questions like a trusted friend, not a moderator. Answers are translated into behavioral instructions, not stored as facts. Personal details are inferred into how Claude should behave differently with that employee, never recorded as private disclosures. ### Draft Review After the 13 questions, Claude writes out a prose draft of the employee's full profile organized under the five section headings. The employee reviews it and can request changes to any section. When they are satisfied, Claude converts the approved prose into the final XML block. ### Final Step The employee pastes the XML into **Settings > Profile > Personal Preferences** in Claude. Done. ### Updating Over Time Employees can run `/personalizer` again anytime. Claude reads their current profile automatically (no paste required) and handles targeted edits, section refreshes, or full rebuilds based on how the employee triggers the skill. *** ## Architecture ``` Employee triggers /personalizer in Claude → SKILL.md (conversation logic + translation rules) → Prose draft presented to employee → Employee reviews and confirms → XML profile prompt generated → Employee pastes into Claude Enterprise Personal Preferences ``` The skill lives entirely in a single SKILL.md file. No external services, no database, no backend. The conversation runs in Claude directly, the profile is generated in Claude directly, and each employee's profile stays private to their own account. *** ## Current Limitations * The skill is deployed as a personal skill for testing. Not yet rolled out org-wide. * The skill does not track adoption. Managers who want to verify rollout need to check with employees directly. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://knowledge.themailworks.com/claude-enterprise/personal-preferences.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.