Update Voice Mmw

<if condition="arguments-provided">
  <action>Look for a matching profile in `<home>/.things/mark-my-words/voices/<argument>.md`.</action>
  <if condition="no-exact-match">Try fuzzy matching against available voice names.</if>
</if>

<if condition="no-arguments-or-no-match">
  <ask-user-question>
    <question>Which voice profile?</question>
    Show each voice's name and description (read from frontmatter).
  </ask-user-question>
</if>

<read path="<home>/.things/mark-my-words/voices/<name>.md" output="profile" />
<action>Display it to the user in full -- frontmatter and all body sections.</action>

<ask-user-question>
  <question>How do you want to update this voice?</question>
  <option>Add samples -- feed in more writing to sharpen the profile</option>
  <option>Manual adjust -- edit specific sections yourself</option>
  <option>Regenerate -- rebuild from scratch using all sources (existing + new)</option>
</ask-user-question>

<if condition="user-add-samples">
  <phase name="add-samples" number="1">
    <action>Follow the same sample-gathering flow as create-voice Step 6 (file paths, paste, or both -- no URLs).</action>

    <action>After gathering new samples, re-analyze all sections in the context of both the existing profile and the new samples. The profile should evolve, not be replaced -- new patterns get added, existing ones get refined or confirmed.</action>

    <action>Update the `sample_sources` list in frontmatter to include the new sources. Update `last_updated` to today's date.</action>

    <constraint>Use Edit to update changed sections. Preserve sections that didn't change.</constraint>
  </phase>
</if>

<if condition="user-manual-adjust">
  <phase name="manual-adjust" number="2">
    <ask-user-question>
      <question>Which section(s) to update?</question>
      <option>1. Tone and Register</option>
      <option>2. Sentence Patterns</option>
      <option>3. Vocabulary and Diction</option>
      <option>4. Rhetorical Habits</option>
      <option>5. Structural Preferences</option>
      <option>6. Things to Avoid</option>
    </ask-user-question>

    <action>For each selected section, show the current content and use AskUserQuestion to get the user's changes. They can describe what to add, remove, or modify.</action>
    <action>Use Edit to apply the changes. Update `last_updated` to today's date.</action>
  </phase>
</if>

<if condition="user-regenerate">
  <phase name="regenerate" number="3">
    <action>Gather new samples using the same flow as create-voice Step 6. Combine with existing `sample_sources` from the current profile.</action>
    <action>Read all available sources (both old file paths and new samples). Re-analyze everything from scratch following the format in <reference name="voice-profile-format"/>.</action>
    <action>Update `last_updated` to today's date. Keep the original `created` date.</action>
    <constraint>Use Write to replace the profile file.</constraint>
  </phase>
</if>

<action>Show the updated profile and ask:</action>

<ask-user-question>
  <question>How does the updated profile look?</question>
  <option>Save it -- looks good</option>
  <option>Keep editing -- I want to change more</option>
  <option>Revert -- go back to the previous version</option>
</ask-user-question>

<if condition="user-save-it">Confirm the profile has been saved.</if>
<if condition="user-keep-editing">Go back to Step 5.</if>
<if condition="user-revert">Write back the original profile content (keep a copy in memory before making changes). Tell the user the profile has been reverted.</if>

<completion-message>
  Tell the user:

  > Voice profile "<name>" updated.

  <action>Check if this voice is set as `default_voice` in `<home>/.things/mark-my-words/preferences.json`.</action>

  <if condition="is-default-voice">

  > This is your default voice -- new posts will use the updated style.

  </if>

  <if condition="not-default-voice">

  > Run `/create-voice-mmw` to create additional profiles, or update `default_voice` in your preferences to make this the default.

  </if>
</completion-message>