memory-init

name	memory-init
description	Initialize the Engram memory system — creates configuration, memory files, and CLAUDE.md integration. Trigger when the user sets up persistent memory for a project, or when `.claude/memory-settings.json` is missing and a memory feature is requested.

Memory Init

Initialize the Engram memory system for the current project.

When to Use

Trigger this skill when:

• The user explicitly asks to set up, initialize, or configure Engram / persistent memory
• A memory-related skill (memory-remember, memory-recall, memory-status) is requested but .claude/memory-settings.json does not exist
• The user re-runs init to upgrade an existing configuration (e.g. to change preset, language, or search_mode)

Workflow

1. Ask Preset Choice

Ask the user which preset they want:

• personal-assistant — Daily life, preferences, habits, personal interactions
• project-assistant — Codebase, architecture, tech decisions, team workflow
• character-companion — Persona, relationship, story progression, emotional bonds
• pair-programmer — AI self-learning, mistake tracking, user preference rules, collaboration patterns

Or offer custom configuration.

2. Ask File Name Customization

Show the default file names and ask if the user wants to change them:

memory_preferences.md    (User/Project/Character preferences)
memory_conversations.md  (Conversation history)
memory_longterm.md       (Long-term memories)

3. Ask Language Preference

• en — English CLAUDE.md section (default)
• zh — Chinese CLAUDE.md section

3.5 Ask Search Mode

• strong — 回覆前預設先用 memory-recall 檢查一次記憶 / Check memories before responding by default
• weak — 判斷話題可能涉及過去時才搜尋 / Search only when a topic may involve past context (default)

4. Create Configuration

Create .claude/memory-settings.json with the chosen settings:

{
  "preset": "<chosen-preset>",
  "files": {
    "preferences": "<preferences-filename>",
    "conversations": "<conversations-filename>",
    "longterm": "<longterm-filename>"
  },
  "language": "<en|zh>",
  "search_mode": "<strong|weak>"
}

5. Copy Preset Templates

Copy the template files from the plugin's templates/presets/<preset>/ directory to the project root. The plugin root is available via ${CLAUDE_PLUGIN_ROOT} or can be found by locating the engram plugin directory.

Read each template file and write it to the project root with the configured file names.

Replace {DATE} placeholders with the current date (run date '+%Y-%m-%d' via Bash to get it).

5c. Append Engram Section to CLAUDE.md

Read the CLAUDE.md section template from the plugin's templates/ directory:

• If language is zh: read from templates/claude-md-section-zh.md
• If language is en: read from templates/claude-md-section-en.md

Replace the following placeholders with the actual configured values:

• {PREFS_FILE} → preferences file name
• {CONVOS_FILE} → conversations file name
• {LONGTERM_FILE} → longterm file name
• {PRESET} → chosen preset name
• {VERSION} → plugin version from .claude-plugin/plugin.json
• {SEARCH_MODE_INSTRUCTION} → resolved based on search_mode and language. Emit as a standalone paragraph (no numbered prefix):
  • strong + en: **Strict search mode**: By default, check memories with the \memory-recall` skill before responding; skip only when the topic is clearly unrelated to history (greetings, pure computation, etc.).`
  • weak + en: **Judgment-based search**: When you judge that a topic may involve past discussions or established preferences, search with the \memory-recall` skill before responding.`
  • strong + zh: **嚴格搜尋模式**：回覆前預設先用 \memory-recall` skill 檢查一次記憶;只有話題明顯與歷史無關（打招呼、純粹計算等)時才略過。`
  • weak + zh: **判斷式搜尋**：當你判斷話題可能涉及過去討論或既有偏好時,先用 \memory-recall` skill 搜尋後再回覆。`

Then apply to CLAUDE.md in the project root:

1. If CLAUDE.md exists and contains <!-- ENGRAM:START -->: Replace everything between <!-- ENGRAM:START --> and <!-- ENGRAM:END --> (inclusive) with the new resolved content
2. If CLAUDE.md exists but has no markers: Append the resolved content to the end of the file (with a blank line separator)
3. If CLAUDE.md does not exist: Create a new CLAUDE.md with only the resolved content

The @{PREFS_FILE} reference inside the section causes Claude Code to auto-load preferences into context on every session start.

6. Upgrade Path (Re-running Init)

If .claude/memory-settings.json already exists:

1. Ask what the user wants to change (preset / language / search_mode / file names)
2. Update .claude/memory-settings.json with the new values
3. Re-generate the CLAUDE.md section (Step 5c) with the new values
4. If file names changed, leave the old files in place and tell the user to move them manually — do not delete existing memory content

7. Optional Customization

Ask the user: "Would you like to add or modify any sections in the preferences template?"

If yes, guide through customization — adding new sections, renaming existing ones, or removing irrelevant ones.

8. Optional Initial Fill

Ask the user: "Would you like to fill in your initial preferences now?"

If yes, walk through each section and populate the preferences file with the user's answers.

9. Output Summary

Display a summary of what was created:

Engram Memory System Initialized!

Preset: <preset>
Language: <language>
Search mode: <search_mode>

Files created/updated:
  - <preferences-file> (preferences — auto-loaded via @ reference in CLAUDE.md)
  - <conversations-file> (conversation history)
  - <longterm-file> (long-term memories)
  - .claude/memory-settings.json (configuration)
  - CLAUDE.md (memory instructions + preferences auto-load)

The memory system is now active. I will:
  - See your preferences every session via the @-reference in CLAUDE.md
  - Save new preferences and decisions as I discover them
  - Record conversation summaries at natural endpoints
  - Search past memories when a topic relates to previous context