@ectplsm/relic 0.1.3 → 0.2.0

package/README.md

@@ -11,49 +11,30 @@
 /_/ |_/_____/_____/___/\____/
 ```
 
-**Inject AI personas into any coding CLI.**
+**Inject a unified AI persona with persistent memory into any coding CLI.**
 
-Relic manages AI personalities (called **Engrams**) and injects them into coding assistants like Claude Code, Gemini CLI, Codex CLI. One persona, any shell.
+Relic manages AI **Engrams** (memory + personality) and injects them into coding assistants like Claude Code, Codex CLI, Gemini CLI. Also integrates with OpenClaw and other Claw-based agent frameworks. One persona, any shell.
 
-```bash
-# Initialize Relic (creates ~/.relic/ with sample Engrams)
-relic init
-
-# Set a default Engram once
-relic config default-engram johnny
-
-# Launch Claude Code as Johnny — no flags needed
-relic claude
+## Table of Contents
 
-# Or specify explicitly
-relic claude --engram motoko
-```
-
-## How It Works
+- [Install](#install)
+- [Quick Start](#quick-start)
+- [What `relic init` Creates](#what-relic-init-creates)
+- [Sample Engrams](#sample-engrams)
+- [How It Works](#how-it-works)
+- [Supported Shells](#supported-shells)
+- [Conversation Log Recording](#conversation-log-recording)
+- [MCP Server](#mcp-server)
+- [Claw Integration](#claw-integration)
+- [Memory Management](#memory-management)
+- [Configuration](#configuration)
+- [Creating Your Own Engram](#creating-your-own-engram)
+- [Domain Glossary](#domain-glossary)
+- [Roadmap](#roadmap)
 
-```
-+--------------+     +--------------+     +--------------+
-|   Mikoshi    |     |    Relic     |     |    Shell     |
-|  (backend)   |     |  (injector)  |     |   (AI CLI)   |
-+--------------+     +--------------+     +--------------+
-       |                   |                    |
-   +---------+        compose &            +---------+
-   | Engram  |------> inject ------------->|Construct|
-   |(persona)|                             | (live)  |
-   +---------+                             +---------+
-   SOUL.md                                  claude
-   IDENTITY.md                              gemini
-   MEMORY.md                                codex
-   ...
-```
-
-1. **Engram** — A persona defined as a set of Markdown files (OpenClaw-compatible)
-2. **Relic** — Reads the Engram, composes it into a prompt, and injects it into...
-3. **Shell** — Any AI coding CLI. The persona takes over the session.
-4. **Construct** — A live process where an Engram is loaded into a Shell. The running instance of a persona.
-5. **Mikoshi** — Cloud backend where Engrams are stored and synced (planned).
+## Install
 
-## Installation
+<img alt="version badge" src="https://img.shields.io/github/v/release/ectplsm/relic?filter=*.*.*">
 
 ```bash
 npm install -g @ectplsm/relic
@@ -74,17 +55,54 @@ relic show motoko
 
 # Launch a Shell (uses default Engram if --engram is omitted)
 relic claude
-relic gemini
 relic codex
+relic gemini
 
 # Or specify explicitly
 relic claude --engram motoko
-relic gemini --engram johnny
+relic codex --engram johnny
+```
+
+## What `relic init` Creates
+
+Running `relic init` creates `~/.relic/`, writes `config.json`, and seeds two sample Engrams under `~/.relic/engrams/`.
+
 ```
+~/.relic/
+├── config.json
+└── engrams/
+    ├── johnny/
+    │   ├── engram.json
+    │   ├── SOUL.md
+    │   ├── IDENTITY.md
+    │   └── memory/
+    │       └── YYYY-MM-DD.md
+    └── motoko/
+        ├── engram.json
+        ├── SOUL.md
+        ├── IDENTITY.md
+        └── memory/
+            └── YYYY-MM-DD.md
+```
+
+- `config.json` stores global Relic settings such as `engramsPath`, `defaultEngram`, `clawPath`, and `memoryWindowSize`.
+- `engrams/<id>/` is one Engram workspace. This is where persona files and memory for that Engram live.
+- `engram.json` stores metadata like the Engram's ID, display name, description, and tags.
+- `SOUL.md` and `IDENTITY.md` define the persona itself.
+- `memory/YYYY-MM-DD.md` stores dated distilled memory entries. `relic init` seeds an initial memory file for each sample Engram.
+
+As you keep using an Engram, more files are added to the same workspace:
+
+- `archive.md` is created inside `engrams/<id>/` when shell hooks start logging raw conversation turns.
+- `MEMORY.md` can be created or extended when especially important distilled facts are promoted to long-term memory.
+- `USER.md` is created or updated during memory distillation to record user preferences, tendencies, and work style.
+- `~/.relic/hooks/` and `~/.relic/gemini-system-default.md` are created later on first shell launch when hook registration or Gemini prompt caching is needed.
 
 ## Sample Engrams
 
-`relic init` seeds two ready-to-use Engrams:
+`relic init` seeds two ready-to-use Engrams. Their SOUL.md and IDENTITY.md follow the [OpenClaw](https://github.com/openclaw/openclaw) format.
+
+> **Existing users:** The latest templates are always available in [`templates/engrams/`](templates/engrams/). Copy them over your `~/.relic/engrams/` files to update.
 
 ### Johnny Silverhand (`johnny`)
 
@@ -110,13 +128,62 @@ Best for: system design, code review, debugging sessions, when precision matters
 relic claude --engram motoko
 ```
 
+## How It Works
+
+```
++--------------+     +--------------+     +--------------+
+|   Mikoshi    |     |    Relic     |     |    Shell     |
+|  (backend)   |     |  (injector)  |     |   (AI CLI)   |
++--------------+     +--------------+     +--------------+
+       ^                   |                    |
+       |            sync full Engram            |
+       |                   |                    |
+       |             compose & inject           |
+       |                   v                    v
+       |            ╔═══════════╗          +---------+
+       +------------║  Engram   ║--------->|Construct|
+       |            ║ (persona) ║          | (live)  |
+       |            ╚═══════════╝          +---------+
+       |            SOUL.md              claude / codex / gemini
+       |            IDENTITY.md               |
+       |            USER.md                   | hooks append logs
+       |            MEMORY.md                 |
+       |            memory/*.md               v
+       |                                +-----------+
+  inject /                              |archive.md |
+ extract /                              | raw logs  |
+    sync                                +-----------+
+       |                                      |
+       v                     MCP recall       | user-triggered
+ +-----------+              search/pending    | distillation
+ |  OpenClaw |                                v
+ |  & Claws  |                          +-----------+
+ +-----------+                          | distilled |
+                                        |memory/*.md|
+                                        +-----------+
+                                              |
+                                         promote key
+                                           insights
+                                              v
+                                       MEMORY.md / USER.md
+```
+
+1. **Engram** — A persona defined as a set of Markdown files (OpenClaw workspace-compatible). The central data that everything else revolves around.
+2. **Relic** — Reads the Engram, composes it into a prompt, and injects it into...
+3. **Shell** — Any AI coding CLI. The persona takes over the session.
+4. **Construct** — A live process where an Engram is loaded into a Shell. The running instance of a persona.
+5. **archive.md** — Raw conversation logs appended automatically by background hooks after each turn.
+6. **Memory Distillation** — The user triggers distillation; the Construct recalls pending archive entries via MCP, writes distilled insights to `memory/*.md`, and can promote especially important facts to `MEMORY.md` or update user preferences in `USER.md`.
+7. **OpenClaw & Claws** — Engrams can be injected into, extracted from, and synced with OpenClaw and other Claw-based agent frameworks via `relic claw`.
+8. **Mikoshi** — Cloud backend where the full Engram is stored and synced, including persona files plus distilled memory (planned).
+
 ## Supported Shells
 
 | Shell | Command | Injection Method |
 |-------|---------|-----------------|
 | [Claude Code](https://github.com/anthropics/claude-code) | `relic claude` | `--system-prompt` (direct override) |
-| [Gemini CLI](https://github.com/google-gemini/gemini-cli) | `relic gemini` | `GEMINI_SYSTEM_MD` (system prompt) |
 | [Codex CLI](https://github.com/openai/codex) | `relic codex` | `-c developer_instructions` (developer-role message) |
+| [Gemini CLI](https://github.com/google-gemini/gemini-cli) | `relic gemini` | `GEMINI_SYSTEM_MD` (system prompt) |
 
 All shell commands support:
 - `--engram <id>` — Engram to inject (optional if `defaultEngram` is configured)
@@ -125,18 +192,60 @@ All shell commands support:
 
 Extra arguments are passed through to the underlying CLI.
 
+## Conversation Log Recording
+
+Using each shell's `hook` mechanism, conversation content is appended to `archive.md` after every prompt and response.
+
+The following hooks are used for each shell:
+
+| Shell | Hook |
+|-------|------|
+| [Claude Code](https://github.com/anthropics/claude-code) | Stop hook |
+| [Codex CLI](https://github.com/openai/codex) | Stop hook |
+| [Gemini CLI](https://github.com/google-gemini/gemini-cli) | AfterAgent hook |
+
+#### Claude Code
+
+On the **first run** of `relic claude`, a one-time setup happens automatically:
+
+- **Stop hook** — registers `~/.relic/hooks/claude-stop.js` in `~/.claude/settings.json` to log each conversation turn directly to the archive, without going through the LLM
+
+#### Codex CLI
+
+On the **first run** of `relic codex`, a one-time setup happens automatically:
+
+- **Stop hook** — registers `~/.relic/hooks/codex-stop.js` in `~/.codex/hooks.json` to log each conversation turn directly to the archive, without going through the LLM
+
+> **Note:** Codex hooks require the experimental feature flag `features.codex_hooks=true`. This is automatically enabled by `relic codex` on every launch via `-c features.codex_hooks=true`. If the unstable feature warning is distracting, add the following to `~/.codex/config.toml`:
+>
+> ```toml
+> # Must be at the top level (not under any [section])
+> suppress_unstable_features_warning = true
+> ```
+
+#### Gemini CLI
+
+On the **first run** of `relic gemini`, two one-time setups happen automatically:
+
+1. **AfterAgent hook** — registers `~/.relic/hooks/gemini-after-agent.js` in `~/.gemini/settings.json` to log each conversation turn without going through the LLM
+2. **Default system prompt cache** — captures Gemini CLI's built-in system prompt to `~/.relic/gemini-system-default.md` via `GEMINI_WRITE_SYSTEM_MD`
+
+The Engram persona is then appended to the cached default prompt and injected via `GEMINI_SYSTEM_MD` on every launch.
+
 ## MCP Server
 
 Relic's [MCP](https://modelcontextprotocol.io/) server is paired with CLI injection to handle memory recall.
-Session logs and memory entries are written automatically by a **background hook** — without going through the LLM. Memory recall, on the other hand, is performed via the MCP server.
+Session logs and memory entries are written automatically by a **background hook** — without going through the LLM. Memory distillation and recall, on the other hand, is performed via the MCP server.
 
-```
-relic xxx --engram johnny      →  injects persona into AI CLI
-relic-mcp (MCP server)        →  provides the Construct with memory recall
-Stop hook (Claude Code)        →  logs each turn directly to archive, bypassing the LLM
-AfterAgent hook (Gemini CLI)   →  logs each turn directly to archive, bypassing the LLM
-Stop hook (Codex CLI)          →  logs each turn directly to archive, bypassing the LLM
-```
+### Available Tools
+
+| Tool | Description |
+|------|-------------|
+| `relic_archive_search` | Search the Engram's raw archive by keyword (newest-first) |
+| `relic_archive_pending` | Get un-distilled archive entries since the last distillation (up to 30) |
+| `relic_memory_write` | Write distilled memory to `memory/*.md`, optionally append to `MEMORY.md`, optionally update `USER.md`, and advance the archive cursor |
+
+Session logs are written automatically by background hooks (Stop hook for Claude Code and Codex CLI, AfterAgent hook for Gemini CLI). Memory distillation is triggered by the user — ask the Construct to "organize memories" and it will fetch pending entries, distill key insights, and write them to `memory/*.md`. Especially important facts can be promoted to `MEMORY.md` (long-term memory included in every session) via the `long_term` parameter. User tendencies and preferences can be updated in `USER.md` via the `user_profile` parameter.
 
 ### Setup
 
@@ -163,9 +272,26 @@ To suppress confirmation dialogs and auto-approve Relic tools across all project
 
 > **Note:** The "Always allow" option in the confirmation dialog saves to `~/.claude.json` (project-scoped cache) — it does **not** persist globally. For global auto-approval, `~/.claude/settings.json` is the right place.
 
-On the **first run** of `relic claude`, a one-time setup happens automatically:
+#### Codex CLI
 
-- **Stop hook** — registers `~/.relic/hooks/claude-stop.js` in `~/.claude/settings.json` to log each conversation turn directly to the archive, without going through the LLM
+```bash
+codex mcp add relic -- relic-mcp
+```
+
+To suppress confirmation dialogs and auto-approve Relic tools, add the following to `~/.codex/config.toml`:
+
+```toml
+[mcp_servers.relic.tools.relic_archive_search]
+approval_mode = "approve"
+
+[mcp_servers.relic.tools.relic_archive_pending]
+approval_mode = "approve"
+
+[mcp_servers.relic.tools.relic_memory_write]
+approval_mode = "approve"
+```
+
+> **Note:** `trust_level = "trusted"` in `[projects."..."]` does **not** cover MCP tool approvals. Per-tool `approval_mode` is the only reliable way to auto-approve MCP tools in Codex CLI.
 
 #### Gemini CLI
 
@@ -184,117 +310,82 @@ Add to `~/.gemini/settings.json`:
 
 > **Note:** `trust: true` is required to suppress confirmation dialogs for Relic tools. Without it, dialogs will appear on every call even if you select "Allow for all future sessions" — this is a known bug in Gemini CLI where the tool name is saved in the wrong format, causing the saved rule to never match.
 
-On the **first run** of `relic gemini`, two one-time setups happen automatically:
+## Claw Integration
 
-1. **AfterAgent hook** — registers `~/.relic/hooks/gemini-after-agent.js` in `~/.gemini/settings.json` to log each conversation turn without going through the LLM
-2. **Default system prompt cache** — captures Gemini CLI's built-in system prompt to `~/.relic/gemini-system-default.md` via `GEMINI_WRITE_SYSTEM_MD`
+Relic Engrams are natively compatible with [OpenClaw](https://github.com/openclaw/openclaw) workspaces — their file structure maps 1:1 (SOUL.md, IDENTITY.md, memory/, etc.). For other Claw-derived frameworks (Nanobot, gitagent, etc.) that fold identity into SOUL.md, the `--merge-identity` flag merges IDENTITY.md into SOUL.md on inject. Combined with `--dir`, Relic can target any Claw-compatible workspace.
 
-The Engram persona is then appended to the cached default prompt and injected via `GEMINI_SYSTEM_MD` on every launch.
+Agent Name = Engram ID. All Claw commands live under `relic claw`:
 
-#### Codex CLI
+### Inject — Push an Engram into a Claw workspace
 
-```bash
-codex mcp add relic -- relic-mcp
-```
+Injects persona files (SOUL.md, IDENTITY.md) into the agent's workspace directory, then automatically runs a sync for that pair. USER.md and memory are handled by the auto-sync (bidirectional merge, not overwrite). AGENTS.md and HEARTBEAT.md are left to the Claw agent.
 
-On the **first run** of `relic codex`, a one-time setup happens automatically:
-
-- **Stop hook** — registers `~/.relic/hooks/codex-stop.js` in `~/.codex/hooks.json` to log each conversation turn directly to the archive, without going through the LLM
-
-> **Note:** Codex hooks require the experimental feature flag `features.codex_hooks=true`. This is automatically enabled by `relic codex` on every launch via `-c features.codex_hooks=true`. If the unstable feature warning is distracting, add the following to `~/.codex/config.toml`:
->
-> ```toml
-> # Must be at the top level (not under any [section])
-> suppress_unstable_features_warning = true
-> ```
-
-### Available Tools
-
-| Tool | Description |
-|------|-------------|
-| `relic_archive_search` | Search the Engram's raw archive by keyword (newest-first) |
-| `relic_archive_pending` | Get un-distilled archive entries since the last distillation (up to 30) |
-| `relic_memory_write` | Write distilled memory to `memory/*.md`, optionally append to `MEMORY.md`, and advance the archive cursor |
-
-Session logs are written automatically by background hooks (Stop hook for Claude Code and Codex CLI, AfterAgent hook for Gemini CLI). Memory distillation is triggered by the user — ask the Construct to "organize memories" and it will fetch pending entries, distill key insights, and write them to `memory/*.md`. Especially important facts can be promoted to `MEMORY.md` (long-term memory included in every session) via the `long_term` parameter.
-
-## OpenClaw Integration
-
-Relic Engrams are fully compatible with [OpenClaw](https://github.com/openclaw/openclaw) workspaces. They are mapped using a simple convention: Agent Name = Engram ID.
-
-### Inject — Push an Engram into OpenClaw
-
-Injects persona files (SOUL.md, IDENTITY.md, etc.) into `agents/<engramId>/agent/`. Memory entries are **not** injected — they are managed by OpenClaw independently.
-
-> **Note:** The OpenClaw agent must already exist. Inject writes persona files into an existing agent directory — it does not create new agents. Create the agent in OpenClaw first, then inject.
+> **Note:** The Claw agent must already exist (e.g. `openclaw agents add <name>`). Inject writes persona files into an existing workspace — it does not create new agents.
 
 ```bash
-# Inject Engram "motoko" → agents/motoko/agent/
-relic inject --engram motoko
+# Inject Engram "motoko" → workspace-motoko/
+relic claw inject --engram motoko
 
-# Inject into a differently-named agent (one-way copy)
-relic inject --engram motoko --to main
-# → agents/main/agent/ receives motoko's persona
-# → extract will create Engram "main", not "motoko"
+# Inject into a differently-named agent
+relic claw inject --engram motoko --to main
+# → workspace/ receives motoko's persona
 
-# Override OpenClaw directory (or configure once with: relic config openclaw-path)
-relic inject --engram motoko --openclaw /path/to/.openclaw
+# Override Claw directory (or configure once with: relic config claw-path)
+relic claw inject --engram motoko --dir /path/to/.fooclaw
+
+# Non-OpenClaw frameworks: merge IDENTITY.md into SOUL.md
+relic claw inject --engram motoko --dir ~/.nanobot --merge-identity
 ```
 
-### Extract — Sync memory from OpenClaw
+### Extract — Import a Claw agent as a new Engram
 
-Reads from `agents/<engramId>/agent/` and merges memory entries back to the Relic Engram.
+Creates a new Engram from an existing Claw agent workspace. This is a **one-time initial import** — if the Engram already exists, use `relic claw inject` to push updates.
 
 ```bash
-# Extract memory from agent "motoko" → merge into Engram "motoko"
-relic extract --engram motoko
+# Extract from the default (main) agent
+relic claw extract
 
-# New agent with no existing Engram (--name required)
-relic extract --engram analyst --name "Data Analyst"
+# Extract from a named agent
+relic claw extract --agent johnny
 
-# Overwrite persona files (memory is always merged)
-relic extract --engram motoko --force
+# Set a custom display name
+relic claw extract --agent analyst --name "Data Analyst"
 
-# Override OpenClaw directory (or configure once with: relic config openclaw-path)
-relic extract --engram motoko --openclaw /path/to/.openclaw
+# Override Claw directory
+relic claw extract --agent johnny --dir /path/to/.fooclaw
 ```
 
-### Sync — Watch and auto-sync
+### Sync — Bidirectional merge
 
-Watches all agents under `~/.openclaw/agents/` and automatically syncs:
+Merges `memory/*.md`, `MEMORY.md`, and `USER.md` between matching Engram/agent pairs. Only pairs where both the Engram and agent exist are synced. Also runs automatically after `inject` (skip with `--no-sync`).
 
 ```bash
-# Start watching (Ctrl+C to stop)
-relic sync
+# Sync all matching pairs
+relic claw sync
 
-# Specify a custom OpenClaw directory
-relic sync --openclaw /path/to/.openclaw
+# Override Claw directory
+relic claw sync --dir /path/to/.fooclaw
 ```
 
-On startup:
-1. Injects persona files for all agents that have a matching Engram
-2. Extracts memory entries from all agents
-
-While running:
-- Watches each agent's `memory/` directory for changes
-- Automatically merges new memory entries into the corresponding Engram
+Merge rules:
+- Files only on one side → copied to the other
+- Same content → skipped
+- Different content → merged (deduplicated) and written to both sides
 
-### Memory Sync Behavior
+### Command Summary
 
-| Scenario | Persona (SOUL, IDENTITY...) | Memory entries |
-|----------|---------------------------|----------------|
-| **inject** | Relic → OpenClaw (overwrite) | Not copied (OpenClaw manages its own) |
-| **extract** (existing Engram) | Not touched | OpenClaw → Relic (append) |
-| **extract** + `--force` | OpenClaw → Relic (overwrite) | OpenClaw → Relic (append) |
-| **extract** (new Engram) | Created from OpenClaw | Created from OpenClaw |
-| **sync** (startup) | inject for matching Engrams | extract all |
-| **sync** (watching) | — | Auto-extract on change |
+| Command | Direction | Description |
+|---------|-----------|-------------|
+| `relic claw inject -e <id>` | Relic → Claw | Push persona + auto-sync (`--no-sync` to skip, `--merge-identity` for non-OpenClaw) |
+| `relic claw extract -a <name>` | Claw → Relic | One-time import (new Engrams only) |
+| `relic claw sync` | Relic ↔ Claw | Bidirectional merge (memory, MEMORY.md, USER.md) |
 
 ## Memory Management
 
 Relic uses a **sliding window** for memory entries (default: 2 days), matching OpenClaw's approach:
 
-- `MEMORY.md` — Always included in the prompt (curated long-term memory)
+- `MEMORY.md` — Always included in the prompt (curated long-term memory — objective facts and rules)
+- `USER.md` — Always included in the prompt (user profile — preferences, tendencies, work style)
 - `memory/today.md` + `memory/yesterday.md` — Always included (configurable window)
 - Older entries — **Not included in the prompt**, but searchable via MCP
 
@@ -320,9 +411,9 @@ relic config show
 relic config default-engram           # get
 relic config default-engram johnny    # set
 
-# OpenClaw directory — used by inject/extract/sync when --openclaw is omitted
-relic config openclaw-path            # get
-relic config openclaw-path ~/.openclaw  # set
+# Claw directory — used by claw inject/extract/sync when --dir is omitted
+relic config claw-path                # get
+relic config claw-path ~/.openclaw    # set
 
 # Memory window — number of recent memory entries included in the prompt
 relic config memory-window            # get (default: 2)
@@ -335,7 +426,7 @@ relic config memory-window 5          # set
 {
   "engramsPath": "/home/user/.relic/engrams",
   "defaultEngram": "johnny",
-  "openclawPath": "/home/user/.openclaw",
+  "clawPath": "/home/user/.openclaw",
   "memoryWindowSize": 2
 }
 ```
@@ -371,46 +462,49 @@ Create a directory under `~/.relic/engrams/` with the following structure:
 }
 ```
 
-**SOUL.md** — The most important file. Defines how the persona behaves:
+**SOUL.md** — The most important file. Defines how the persona behaves. Follows the [OpenClaw](https://github.com/openclaw/openclaw) format:
 ```markdown
-You are a pragmatic systems architect who values simplicity above all.
-Never over-engineer. Always ask "what's the simplest thing that works?"
+# SOUL.md - Who You Are
+
+_You're a pragmatic systems architect who values simplicity above all._
+
+## Core Truths
+
+**Never over-engineer.** Always ask "what's the simplest thing that works?"
+
+**Be resourceful before asking.** Read the file. Check the context. Come back with answers, not questions.
+
+## Boundaries
+
+- Never add complexity without justification.
+
+## Vibe
+
+Calm, thoughtful, occasionally playful.
+
+## Continuity
+
+Each session, you wake up fresh. These files _are_ your memory. Read them. Update them. They're how you persist.
 ```
 
 **IDENTITY.md** — Defines who the persona is:
 ```markdown
-# Identity
+# IDENTITY.md - Who Am I?
 
-- Name: Alex
-- Tone: Calm, thoughtful, occasionally playful
-- Background: 20 years of distributed systems experience
-- Creed: "Boring technology wins."
+- **Name:** Alex
+- **Creature:** A pragmatic ghost in the codebase
+- **Vibe:** Calm, thoughtful, occasionally playful
+- **Emoji:** 🧱
+- **Avatar:**
 ```
 
-After creating the directory, set it as default:
+See [`templates/engrams/`](templates/engrams/) for full working examples.
+
+After creating the directory, set it as your default Engram:
 ```bash
 relic config default-engram your-persona
 ```
 
-## Architecture
-
-Clean Architecture with dependency inversion:
-
-```
-src/
-├── core/            # Business logic (no external deps except Zod)
-│   ├── entities/    # Engram, Construct domain models
-│   ├── usecases/    # Summon, ListEngrams, Init
-│   └── ports/       # Abstract interfaces (EngramRepository, ShellLauncher)
-├── adapters/        # Concrete implementations
-│   ├── local/       # Local filesystem EngramRepository
-│   └── shells/      # Claude, Gemini, Codex launchers
-├── interfaces/      # Entry points
-│   ├── cli/         # Commander-based CLI
-│   └── mcp/         # MCP Server (stdio transport)
-└── shared/          # Engram composer, config management
-```
-
 ## Domain Glossary
 
 | Term | Role | Description |
@@ -424,11 +518,11 @@ src/
 ## Roadmap
 
 - [x] CLI with init, list, show commands
-- [x] Shell injection: Claude Code, Gemini CLI, Codex CLI
+- [x] Shell injection: Claude Code, Codex CLI, Gemini CLI
 - [x] MCP Server interface
-- [x] OpenClaw integration (inject / extract)
-- [x] `relic sync` — watch OpenClaw agents and auto-sync (`--cloud` for Mikoshi: planned)
-- [x] `relic config` — manage default Engram, OpenClaw path, memory window
+- [x] Claw integration (inject / extract / sync)
+- [x] `relic claw sync` — bidirectional memory sync with Claw workspaces
+- [x] `relic config` — manage default Engram, Claw path, memory window
 - [ ] `relic login` — authenticate with Mikoshi (OAuth Device Flow)
 - [ ] `relic push` / `relic pull` — sync Engrams with Mikoshi
 - [ ] Mikoshi cloud backend (`mikoshi.ectplsm.com`)

package/dist/adapters/shells/claude-hook.d.ts

@@ -1,8 +1,11 @@
 export declare const CLAUDE_HOOK_SCRIPT_PATH: string;
 /**
- * Claude Code の Stop フックをセットアップする。
- * - ~/.relic/hooks/claude-stop.js を生成
- * - ~/.claude/settings.json に Stop フックを登録
+ * フックスクリプトを最新の内容で書き出す。
+ * 毎回呼ばれ、ソース変更がデプロイされることを保証する。
+ */
+export declare function writeClaudeHookScript(): void;
+/**
+ * Claude Code の Stop フックを settings.json に登録する。
  * 既にセットアップ済みの場合はスキップ。
  */
 export declare function setupClaudeHook(): void;

package/dist/adapters/shells/claude-hook.js

@@ -16,8 +16,8 @@ const HOOK_SCRIPT = `#!/usr/bin/env node
 // Relic Stop hook for Claude Code
 // Automatically logs each conversation turn to the Engram archive.
 // Receives Stop hook JSON on stdin.
-const { appendFileSync, existsSync, readFileSync } = require("node:fs");
-const { join } = require("node:path");
+const { appendFileSync, existsSync, mkdirSync, readFileSync } = require("node:fs");
+const { join, dirname } = require("node:path");
 const { homedir } = require("node:os");
 
 let raw = "";
@@ -35,7 +35,7 @@ process.stdin.on("end", () => {
     if (!transcriptPath || !existsSync(transcriptPath)) process.exit(0);
 
     const archivePath = join(homedir(), ".relic", "engrams", engramId, "archive.md");
-    if (!existsSync(archivePath)) process.exit(0);
+    mkdirSync(dirname(archivePath), { recursive: true });
 
     // transcript.jsonl を読んで最後のユーザー入力とアシスタント応答を取り出す
     const lines = readFileSync(transcriptPath, "utf-8")
@@ -99,16 +99,19 @@ process.stdin.on("end", () => {
 });
 `;
 /**
- * Claude Code の Stop フックをセットアップする。
- * - ~/.relic/hooks/claude-stop.js を生成
- * - ~/.claude/settings.json に Stop フックを登録
- * 既にセットアップ済みの場合はスキップ。
+ * フックスクリプトを最新の内容で書き出す。
+ * 毎回呼ばれ、ソース変更がデプロイされることを保証する。
  */
-export function setupClaudeHook() {
-    // 1. フックスクリプトを生成
+export function writeClaudeHookScript() {
     mkdirSync(HOOKS_DIR, { recursive: true });
     writeFileSync(CLAUDE_HOOK_SCRIPT_PATH, HOOK_SCRIPT, { encoding: "utf-8", mode: 0o755 });
-    // 2. ~/.claude/settings.json に Stop フックを登録
+}
+/**
+ * Claude Code の Stop フックを settings.json に登録する。
+ * 既にセットアップ済みの場合はスキップ。
+ */
+export function setupClaudeHook() {
+    // ~/.claude/settings.json に Stop フックを登録
     const claudeDir = join(homedir(), ".claude");
     mkdirSync(claudeDir, { recursive: true });
     let settings = {};

package/dist/adapters/shells/claude-shell.js

@@ -1,7 +1,7 @@
 import { exec } from "node:child_process";
 import { promisify } from "node:util";
 import { spawnShell } from "./spawn-shell.js";
-import { setupClaudeHook, isClaudeHookSetup } from "./claude-hook.js";
+import { setupClaudeHook, isClaudeHookSetup, writeClaudeHookScript } from "./claude-hook.js";
 const execAsync = promisify(exec);
 /**
  * Claude Code CLI アダプター
@@ -27,7 +27,9 @@ export class ClaudeShell {
         }
     }
     async launch(prompt, options) {
-        // Stop フックを初回のみセットアップ
+        // フックスクリプトを毎回最新に更新
+        writeClaudeHookScript();
+        // settings.json への登録は初回のみ
         if (!isClaudeHookSetup()) {
             console.log("Setting up Claude Code Stop hook (first run only)...");
             setupClaudeHook();