mempalace-code 1.3.0__tar.gz → 1.4.1__tar.gz

mempalace_code-1.4.1/.github/workflows/ci.yml

@@ -0,0 +1,29 @@
+name: Tests
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v6
+        with:
+          python-version: "3.13"
+      - run: pip install -e ".[dev,treesitter]"
+      - run: python3 -m pytest tests/ -v -m "not needs_network"
+
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v6
+        with:
+          python-version: "3.13"
+      - run: pip install ruff
+      - run: ruff check .
+      - run: ruff format --check .

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/.github/workflows/publish.yml

@@ -11,13 +11,13 @@ jobs:
   build:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v4
-      - uses: actions/setup-python@v5
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v6
         with:
           python-version: "3.11"
       - run: pip install build
       - run: python -m build
-      - uses: actions/upload-artifact@v4
+      - uses: actions/upload-artifact@v5
         with:
           name: dist
           path: dist/
@@ -29,7 +29,7 @@ jobs:
     permissions:
       id-token: write
     steps:
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         with:
           name: dist
           path: dist/

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/CHANGELOG.md

@@ -1,5 +1,24 @@
 # Changelog
 
+## v1.4.1 — 2026-04-20
+
+### Changed
+- **Docs: hooks are legacy** — MCP tools + usage rules are now the recommended approach for all agents (Claude Code, Codex, Cursor); hooks demoted to optional Claude Code-only extra
+- **Docs: unified saving story** — README, AGENT_INSTALL, and hooks README all consistently describe watcher for code mining + MCP for conversation context
+
+## v1.4.0 — 2026-04-19
+
+### Added
+- **Watcher quiet mode** — re-mines suppress verbose output; only logs a one-line summary when drawers are actually filed; no-op commits produce zero log noise; optimize skipped on empty batches
+- **Per-project `bin/` skip** — `bin/` no longer globally skipped; only excluded when .NET project markers (`.csproj`, `.sln`, `.fsproj`, `.vbproj`) are present (MINE-BIN-SKIP-DIRS)
+- **Kotlin nested generic receiver** — `fun <T> List<Pair<K,V>>.ext()` now parsed correctly (MINE-KOTLIN-GENERIC-RECEIVER-NESTED)
+- `mine()` now returns stats dict (`files_processed`, `drawers_filed`, `elapsed_secs`)
+
+### Fixed
+- **Watcher on-commit detection** — `watchfiles.DefaultFilter` ignores `.git/` by default; on-commit mode now passes `watch_filter=None` so `.git/refs/heads/` changes are detected
+- **Watcher log buffering** — flush Python stdout/stderr before restoring file descriptors to prevent mine() output leaking to real stdout
+- **HuggingFace/safetensors noise** — suppress BertModel LOAD REPORT and progress bars via OS fd-level redirect during model init
+
 ## v1.3.0 — 2026-04-19
 
 First-class C#/.NET support — delivers [rergards/mempalace-code#1](https://github.com/rergards/mempalace-code/issues/1) in full.

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/CLAUDE.md

@@ -2,7 +2,7 @@
 
 ## Stack
 
-- **Python** 3.9+ (supports 3.9–3.12)
+- **Python** 3.11+ (supports 3.11–3.14)
 - **Storage**: LanceDB (core, crash-safe vector DB — no server required)
 - **Embeddings**: sentence-transformers (local, no API key)
 - **Config**: PyYAML
@@ -72,7 +72,8 @@ Line length: 100. Target: py39. Quote style: double.
 | `layers.py` | Tiered context loading — L0/L1/L2/L3 wake-up layers for local models |
 | `palace_graph.py` | Graph traversal and tunnel detection across wings/rooms |
 | `mcp_server.py` | MCP server — exposes palace tools to Claude Code and other MCP clients |
-| `cli.py` | `mempalace` CLI entry point — init, mine, search, health, repair, backup |
+| `watcher.py` | File watcher — `watch_and_mine`, `watch_all`, launchd/cron schedule rendering |
+| `cli.py` | `mempalace` CLI entry point — init, mine, mine-all, watch, search, health, repair, backup |
 
 ## Architecture Principles
 

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mempalace-code
-Version: 1.3.0
+Version: 1.4.1
 Summary: Developer memory tool — mine codebases and conversations into a LanceDB-backed searchable palace. No API key required.
 Project-URL: Homepage, https://github.com/rergards/mempalace-code
 Project-URL: Repository, https://github.com/rergards/mempalace-code
@@ -16,7 +16,6 @@ Classifier: Development Status :: 4 - Beta
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
@@ -24,7 +23,7 @@ Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Topic :: Utilities
-Requires-Python: >=3.9
+Requires-Python: >=3.11
 Requires-Dist: lancedb>=0.17
 Requires-Dist: pyyaml>=6.0
 Requires-Dist: sentence-transformers>=2.2
@@ -33,15 +32,15 @@ Requires-Dist: chromadb<2,>=0.5.0; extra == 'chroma'
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == 'dev'
 Requires-Dist: ruff>=0.4.0; extra == 'dev'
+Requires-Dist: watchfiles>=1.0; extra == 'dev'
 Provides-Extra: spellcheck
 Requires-Dist: autocorrect>=2.0; extra == 'spellcheck'
 Provides-Extra: treesitter
 Requires-Dist: tree-sitter-go<0.26,>=0.23; extra == 'treesitter'
-Requires-Dist: tree-sitter-python<0.26,>=0.23; (python_version >= '3.10') and extra == 'treesitter'
+Requires-Dist: tree-sitter-python<0.26,>=0.23; extra == 'treesitter'
 Requires-Dist: tree-sitter-rust<0.26,>=0.23; extra == 'treesitter'
 Requires-Dist: tree-sitter-typescript<0.26,>=0.23; extra == 'treesitter'
-Requires-Dist: tree-sitter<0.24,>=0.22; (python_version < '3.10') and extra == 'treesitter'
-Requires-Dist: tree-sitter<0.26,>=0.22; (python_version >= '3.10') and extra == 'treesitter'
+Requires-Dist: tree-sitter<0.26,>=0.22; extra == 'treesitter'
 Provides-Extra: watch
 Requires-Dist: watchfiles>=1.0; extra == 'watch'
 Description-Content-Type: text/markdown
@@ -77,7 +76,7 @@ No cloud. No API keys. No subscription. Nothing leaves your machine.
 <tr>
 <td align="center"><strong>595x Token Savings</strong><br><sub>measured peak · median 80x<br><a href="docs/BENCH_TOKEN_DELTA.md">scales with project size</a></sub></td>
 <td align="center"><strong>Cross-Project Tunnels</strong><br><sub>Search <code>auth</code> in one project<br>find it everywhere</sub></td>
-<td align="center"><strong>1002 Tests · $0 Cost</strong><br><sub>Every feature acceptance-gated<br>fully offline after install</sub></td>
+<td align="center"><strong>1008 Tests · $0 Cost</strong><br><sub>Every feature acceptance-gated<br>fully offline after install</sub></td>
 </tr>
 </table>
 
@@ -88,7 +87,13 @@ No cloud. No API keys. No subscription. Nothing leaves your machine.
 ## Quick Start
 
 ```bash
-pip install mempalace-code
+uv tool install mempalace-code        # recommended (fast, Rust-based)
+# or
+pipx install mempalace-code           # alternative
+# or
+pip install mempalace-code            # into current environment
+# or
+uvx --from mempalace-code mempalace --help  # try without installing
 ```
 
 Then ask your AI to read [`docs/AGENT_INSTALL.md`](docs/AGENT_INSTALL.md) — it will handle setup, MCP wiring, prompt injection, and verification automatically.
@@ -102,6 +107,11 @@ mempalace mine ~/projects/myapp       # index your codebase
 claude mcp add mempalace -- python -m mempalace.mcp_server  # connect to Claude Code
 ```
 
+**Optional: auto-sync on commit** (requires `[watch]` extra — see [Auto-Watch](#auto-watch)):
+```bash
+mempalace watch ~/projects/           # re-mines on every commit, zero noise
+```
+
 This makes the 27 tools available to your AI. For proactive search and storage (without you asking), you'll also need to add usage rules to your `CLAUDE.md` — see [`docs/AGENT_INSTALL.md`](docs/AGENT_INSTALL.md) Section 7.
 
 </details>
@@ -175,10 +185,40 @@ Tree-sitter is optional (`pip install "mempalace-code[treesitter]"`). Without it
 mempalace mine ~/projects/myapp                  # all supported file types
 mempalace mine ~/projects/myapp --wing myapp     # tag with a specific wing
 mempalace mine ~/chats/ --mode convos            # mine conversation exports
+mempalace mine-all ~/projects/                   # batch mine all projects in a directory
 ```
 
 Mining is **incremental** by default — content-hash based, only changed files are re-chunked. Use `--full` to force a rebuild.
 
+### Auto-Watch
+
+Keep your palace in sync automatically. By default, watches `.git/refs/heads/` and re-mines only on **commit** — no noise from work-in-progress saves. Handles multiple branches and worktrees.
+
+Requires the `watch` extra:
+```bash
+uv tool install "mempalace-code[watch]"   # or: pipx install "mempalace-code[watch]"
+```
+
+Already installed without it? Add watchfiles:
+```bash
+uv tool inject mempalace-code watchfiles  # or: pipx inject mempalace-code watchfiles
+```
+
+```bash
+mempalace watch ~/projects/                      # watch all projects (on commit, default)
+mempalace watch ~/projects/ --on-save            # watch all file saves instead (noisier)
+mempalace watch ~/projects/ schedule             # print launchd/cron snippet for daemon
+```
+
+**Install as persistent daemon (macOS):**
+
+```bash
+mempalace watch ~/projects/ schedule > ~/Library/LaunchAgents/com.mempalace.watch.plist
+launchctl load ~/Library/LaunchAgents/com.mempalace.watch.plist
+```
+
+Starts at login, restarts if crashed. Logs to `/tmp/mempalace-watch.log`.
+
 ---
 
 ### The Palace
@@ -421,7 +461,7 @@ This is a code-first fork of [milla-jovovich/mempalace](https://github.com/milla
 | ChromaDB — [silently deletes data on version bump](https://github.com/milla-jovovich/mempalace/issues/469) | LanceDB — crash-safe Arrow storage, no version-cliff |
 | "No internet after install" — [false](https://github.com/milla-jovovich/mempalace/issues/524) | `mempalace init` downloads model explicitly; fully offline after |
 | "100% R@5" — [unverifiable](https://github.com/milla-jovovich/mempalace/issues/27) | Number removed. Methodology caveats documented |
-| ~30% test coverage | 1002 tests, every feature acceptance-gated |
+| ~30% test coverage | 1008 tests, every feature acceptance-gated |
 | No backup, no recovery | `backup` / `restore` / `export` / `import` |
 | No incremental mining | Content-hash incremental: only changed files re-chunked |
 | No code-search | `code_search` — filter by language, symbol, glob |
@@ -496,6 +536,10 @@ mempalace mine <dir> --full                       # force full rebuild
 mempalace mine <dir> --watch                      # auto-incremental on file changes
 mempalace mine-all <parent-dir>                   # batch mine all projects in a directory
 
+# Watch (multi-project auto-sync)
+mempalace watch <parent-dir>                      # watch all initialized projects
+mempalace watch <parent-dir> schedule             # print launchd/cron daemon snippet
+
 # Search
 mempalace search "query"                          # search everything
 mempalace search "query" --wing myapp             # scoped to wing
@@ -523,21 +567,15 @@ mempalace fetch-model                             # pre-download for offline use
 </details>
 
 <details>
-<summary><strong>Auto-Save Hooks</strong></summary>
+<summary><strong>Saving Conversation Context</strong></summary>
 
-Two Claude Code hooks for automatic memory saving:
+Code mining is automatic via `mempalace watch-all`. For conversation context (decisions, discussions, debugging notes), the AI uses MCP tools directly — works with **any agent** (Claude Code, Codex, Cursor, etc.):
 
-- **Stop Hook** — after each response, saves topics, decisions, and code changes
-- **PreCompact Hook** — emergency save before context compression
+1. Wire the MCP server (see [install docs](docs/AGENT_INSTALL.md))
+2. Add usage rules to your agent's instructions (CLAUDE.md, system prompt, etc.)
+3. The agent calls `mempalace_add_drawer` and `mempalace_diary_write` during sessions
 
-```json
-{
-  "hooks": {
-    "Stop": [{"matcher": "", "hooks": [{"type": "command", "command": "/path/to/mempalace/hooks/mempal_save_hook.sh"}]}],
-    "PreCompact": [{"matcher": "", "hooks": [{"type": "command", "command": "/path/to/mempalace/hooks/mempal_precompact_hook.sh"}]}]
-  }
-}
-```
+> **Legacy:** Claude Code also supports optional [auto-save hooks](hooks/README.md) that remind the AI to save at fixed intervals. These are redundant if MCP + usage rules are set up.
 
 </details>
 
@@ -557,9 +595,9 @@ mempalace/
 │   ├── palace_graph.py     ← room navigation graph
 │   └── layers.py           ← 4-layer memory stack
 ├── benchmarks/             ← reproducible benchmark runners
-├── hooks/                  ← Claude Code auto-save hooks
+├── hooks/                  ← Claude Code auto-save hooks (legacy, optional)
 ├── examples/               ← usage examples
-└── tests/                  ← 1002 tests
+└── tests/                  ← 1008 tests
 ```
 
 </details>
@@ -579,7 +617,7 @@ python -m pytest tests/ -x -q    # full suite, all local, no network
 Apache 2.0 — see [LICENSE](LICENSE) and [NOTICE](NOTICE).
 
 <!-- Link Definitions -->
-[version-shield]: https://img.shields.io/badge/version-1.3.0-4dc9f6?style=flat-square&labelColor=0a0e14
+[version-shield]: https://img.shields.io/badge/version-1.4.0-4dc9f6?style=flat-square&labelColor=0a0e14
 [release-link]: https://github.com/rergards/mempalace-code/releases
 [python-shield]: https://img.shields.io/badge/python-3.9+-7dd8f8?style=flat-square&labelColor=0a0e14&logo=python&logoColor=7dd8f8
 [python-link]: https://www.python.org/

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/README.md

@@ -29,7 +29,7 @@ No cloud. No API keys. No subscription. Nothing leaves your machine.
 <tr>
 <td align="center"><strong>595x Token Savings</strong><br><sub>measured peak · median 80x<br><a href="docs/BENCH_TOKEN_DELTA.md">scales with project size</a></sub></td>
 <td align="center"><strong>Cross-Project Tunnels</strong><br><sub>Search <code>auth</code> in one project<br>find it everywhere</sub></td>
-<td align="center"><strong>1002 Tests · $0 Cost</strong><br><sub>Every feature acceptance-gated<br>fully offline after install</sub></td>
+<td align="center"><strong>1008 Tests · $0 Cost</strong><br><sub>Every feature acceptance-gated<br>fully offline after install</sub></td>
 </tr>
 </table>
 
@@ -40,7 +40,13 @@ No cloud. No API keys. No subscription. Nothing leaves your machine.
 ## Quick Start
 
 ```bash
-pip install mempalace-code
+uv tool install mempalace-code        # recommended (fast, Rust-based)
+# or
+pipx install mempalace-code           # alternative
+# or
+pip install mempalace-code            # into current environment
+# or
+uvx --from mempalace-code mempalace --help  # try without installing
 ```
 
 Then ask your AI to read [`docs/AGENT_INSTALL.md`](docs/AGENT_INSTALL.md) — it will handle setup, MCP wiring, prompt injection, and verification automatically.
@@ -54,6 +60,11 @@ mempalace mine ~/projects/myapp       # index your codebase
 claude mcp add mempalace -- python -m mempalace.mcp_server  # connect to Claude Code
 ```
 
+**Optional: auto-sync on commit** (requires `[watch]` extra — see [Auto-Watch](#auto-watch)):
+```bash
+mempalace watch ~/projects/           # re-mines on every commit, zero noise
+```
+
 This makes the 27 tools available to your AI. For proactive search and storage (without you asking), you'll also need to add usage rules to your `CLAUDE.md` — see [`docs/AGENT_INSTALL.md`](docs/AGENT_INSTALL.md) Section 7.
 
 </details>
@@ -127,10 +138,40 @@ Tree-sitter is optional (`pip install "mempalace-code[treesitter]"`). Without it
 mempalace mine ~/projects/myapp                  # all supported file types
 mempalace mine ~/projects/myapp --wing myapp     # tag with a specific wing
 mempalace mine ~/chats/ --mode convos            # mine conversation exports
+mempalace mine-all ~/projects/                   # batch mine all projects in a directory
 ```
 
 Mining is **incremental** by default — content-hash based, only changed files are re-chunked. Use `--full` to force a rebuild.
 
+### Auto-Watch
+
+Keep your palace in sync automatically. By default, watches `.git/refs/heads/` and re-mines only on **commit** — no noise from work-in-progress saves. Handles multiple branches and worktrees.
+
+Requires the `watch` extra:
+```bash
+uv tool install "mempalace-code[watch]"   # or: pipx install "mempalace-code[watch]"
+```
+
+Already installed without it? Add watchfiles:
+```bash
+uv tool inject mempalace-code watchfiles  # or: pipx inject mempalace-code watchfiles
+```
+
+```bash
+mempalace watch ~/projects/                      # watch all projects (on commit, default)
+mempalace watch ~/projects/ --on-save            # watch all file saves instead (noisier)
+mempalace watch ~/projects/ schedule             # print launchd/cron snippet for daemon
+```
+
+**Install as persistent daemon (macOS):**
+
+```bash
+mempalace watch ~/projects/ schedule > ~/Library/LaunchAgents/com.mempalace.watch.plist
+launchctl load ~/Library/LaunchAgents/com.mempalace.watch.plist
+```
+
+Starts at login, restarts if crashed. Logs to `/tmp/mempalace-watch.log`.
+
 ---
 
 ### The Palace
@@ -373,7 +414,7 @@ This is a code-first fork of [milla-jovovich/mempalace](https://github.com/milla
 | ChromaDB — [silently deletes data on version bump](https://github.com/milla-jovovich/mempalace/issues/469) | LanceDB — crash-safe Arrow storage, no version-cliff |
 | "No internet after install" — [false](https://github.com/milla-jovovich/mempalace/issues/524) | `mempalace init` downloads model explicitly; fully offline after |
 | "100% R@5" — [unverifiable](https://github.com/milla-jovovich/mempalace/issues/27) | Number removed. Methodology caveats documented |
-| ~30% test coverage | 1002 tests, every feature acceptance-gated |
+| ~30% test coverage | 1008 tests, every feature acceptance-gated |
 | No backup, no recovery | `backup` / `restore` / `export` / `import` |
 | No incremental mining | Content-hash incremental: only changed files re-chunked |
 | No code-search | `code_search` — filter by language, symbol, glob |
@@ -448,6 +489,10 @@ mempalace mine <dir> --full                       # force full rebuild
 mempalace mine <dir> --watch                      # auto-incremental on file changes
 mempalace mine-all <parent-dir>                   # batch mine all projects in a directory
 
+# Watch (multi-project auto-sync)
+mempalace watch <parent-dir>                      # watch all initialized projects
+mempalace watch <parent-dir> schedule             # print launchd/cron daemon snippet
+
 # Search
 mempalace search "query"                          # search everything
 mempalace search "query" --wing myapp             # scoped to wing
@@ -475,21 +520,15 @@ mempalace fetch-model                             # pre-download for offline use
 </details>
 
 <details>
-<summary><strong>Auto-Save Hooks</strong></summary>
+<summary><strong>Saving Conversation Context</strong></summary>
 
-Two Claude Code hooks for automatic memory saving:
+Code mining is automatic via `mempalace watch-all`. For conversation context (decisions, discussions, debugging notes), the AI uses MCP tools directly — works with **any agent** (Claude Code, Codex, Cursor, etc.):
 
-- **Stop Hook** — after each response, saves topics, decisions, and code changes
-- **PreCompact Hook** — emergency save before context compression
+1. Wire the MCP server (see [install docs](docs/AGENT_INSTALL.md))
+2. Add usage rules to your agent's instructions (CLAUDE.md, system prompt, etc.)
+3. The agent calls `mempalace_add_drawer` and `mempalace_diary_write` during sessions
 
-```json
-{
-  "hooks": {
-    "Stop": [{"matcher": "", "hooks": [{"type": "command", "command": "/path/to/mempalace/hooks/mempal_save_hook.sh"}]}],
-    "PreCompact": [{"matcher": "", "hooks": [{"type": "command", "command": "/path/to/mempalace/hooks/mempal_precompact_hook.sh"}]}]
-  }
-}
-```
+> **Legacy:** Claude Code also supports optional [auto-save hooks](hooks/README.md) that remind the AI to save at fixed intervals. These are redundant if MCP + usage rules are set up.
 
 </details>
 
@@ -509,9 +548,9 @@ mempalace/
 │   ├── palace_graph.py     ← room navigation graph
 │   └── layers.py           ← 4-layer memory stack
 ├── benchmarks/             ← reproducible benchmark runners
-├── hooks/                  ← Claude Code auto-save hooks
+├── hooks/                  ← Claude Code auto-save hooks (legacy, optional)
 ├── examples/               ← usage examples
-└── tests/                  ← 1002 tests
+└── tests/                  ← 1008 tests
 ```
 
 </details>
@@ -531,7 +570,7 @@ python -m pytest tests/ -x -q    # full suite, all local, no network
 Apache 2.0 — see [LICENSE](LICENSE) and [NOTICE](NOTICE).
 
 <!-- Link Definitions -->
-[version-shield]: https://img.shields.io/badge/version-1.3.0-4dc9f6?style=flat-square&labelColor=0a0e14
+[version-shield]: https://img.shields.io/badge/version-1.4.0-4dc9f6?style=flat-square&labelColor=0a0e14
 [release-link]: https://github.com/rergards/mempalace-code/releases
 [python-shield]: https://img.shields.io/badge/python-3.9+-7dd8f8?style=flat-square&labelColor=0a0e14&logo=python&logoColor=7dd8f8
 [python-link]: https://www.python.org/

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/docs/AGENT_INSTALL.md

@@ -29,8 +29,8 @@ Parse `major.minor` from stdout (e.g. `Python 3.11.4` → `3.11`).
 
 **Pass →** `major >= 3` and `minor >= 9`. Record Python binary as `PYTHON=python3`. Continue to Step 1.2.
 
-**Fail →** Python is absent or version < 3.9. **ASK HUMAN:** "Python 3.9 or later is required but was not found (or is too old). Please install Python 3.9+ and re-run this script. Reply `ready` when done."
-Wait for `ready`. Re-run Step 1.1. If still failing after one retry, halt and report: "Cannot proceed — Python 3.9+ is required."
+**Fail →** Python is absent or version < 3.11. **ASK HUMAN:** "Python 3.11 or later is required but was not found (or is too old). Please install Python 3.11+ and re-run this script. Reply `ready` when done."
+Wait for `ready`. Re-run Step 1.1. If still failing after one retry, halt and report: "Cannot proceed — Python 3.11+ is required."
 
 ---
 
@@ -109,10 +109,10 @@ Ask all five questions before acting. Record answers; they parameterize Sections
 
 **Condition:** `ALREADY_INSTALLED=false`
 
-**ASK HUMAN:** "I can install mempalace at the user level (recommended: `pipx`, isolated from other packages) or as a project dependency in the current virtual environment (`uv`/`pip`). Reply `user` for pipx or `project` for the current venv."
+**ASK HUMAN:** "I can install mempalace at the user level (recommended: `uv tool install` or `pipx`, isolated from other packages) or as a project dependency in the current virtual environment (`pip`). Reply `user` for isolated install or `project` for the current venv."
 
 **Parse response:**
-- `user` → Set `INSTALL_METHOD=user`. (If `HAS_PIPX=false`, note pipx must be installed first — see Step 3.1 fallback.)
+- `user` → Set `INSTALL_METHOD=user`. Prefer `uv tool install` if `HAS_UV=true`, else fall back to `pipx`.
 - `project` → Set `INSTALL_METHOD=project`.
 - Anything else → Repeat the question once. If still unclear, default to `user` and inform the human.
 
@@ -189,7 +189,7 @@ curl -fsSL https://raw.githubusercontent.com/rergards/mempalace-code/main/script
 
 **Pass →** Script exits 0 and prints `Done. mempalace <version> is ready.` Skip to Section 4.
 
-**Fail →** Script exits non-zero. Read the error output. Common causes: no Python 3.9+, no `venv` module (`apt install python3-venv`). Fix and re-run, or fall through to manual Steps 3.1–3.3.
+**Fail →** Script exits non-zero. Read the error output. Common causes: no Python 3.11+, no `venv` module (`apt install python3-venv`). Fix and re-run, or fall through to manual Steps 3.1–3.3.
 
 ---
 
@@ -215,7 +215,7 @@ If that succeeds, re-run `pipx install mempalace-code`. If pipx install fails, f
 
 ---
 
-### Step 3.2: Install via uv (INSTALL_METHOD=project or pipx unavailable, HAS_UV=true)
+### Step 3.2: Install via uv (INSTALL_METHOD=user and HAS_UV=true, or pipx unavailable)
 
 **Check:**
 ```bash
@@ -224,7 +224,7 @@ command -v uv
 
 **Pass →** Run:
 ```bash
-uv pip install mempalace-code
+uv tool install mempalace-code
 ```
 Exit code 0 = success. Continue to Step 3.4.
 
@@ -491,14 +491,17 @@ mkdir -p ~/.codex
 
 ---
 
-### Step 5.3: Auto-save hooks (optional, recommended)
+### Step 5.3: Auto-save for conversation context
 
-Auto-save hooks are separate from MCP wiring. They trigger automatic memory saves during AI sessions. Configuration is documented in [`hooks/README.md`](../hooks/README.md). Do NOT inline hook setup here — reference the README and let the human decide.
+Code mining is handled by the watcher (`mempalace watch-all`) and works with any client. Conversation context (decisions, discussions, debugging notes) is saved via **MCP tools + usage rules** — this works identically across all agents.
 
-**ASK HUMAN:** "Would you like to set up auto-save hooks (these automatically save memories during AI sessions — see `hooks/README.md`)? Reply `yes` or `no`."
+The recommended approach for **all agents** (Claude Code, Codex, Cursor, etc.):
+1. Wire the MCP server (Steps 5.1/5.2) so the agent can call `mempalace_add_drawer` and `mempalace_diary_write`.
+2. Add usage rules to the agent's instructions (Section 7) so it knows when to save.
 
-- `yes` → Refer human to `hooks/README.md` for configuration instructions. Do not proceed autonomously.
-- `no` → Skip.
+That's it. No hooks needed.
+
+> **Legacy: Claude Code auto-save hooks.** Claude Code also supports optional bash hooks that fire on Stop/PreCompact events and remind the AI to save at fixed intervals. These are redundant if you have MCP + usage rules set up, but are documented in [`hooks/README.md`](../hooks/README.md) for users who want belt-and-suspenders.
 
 ---
 

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/docs/BACKLOG-archived.yaml

@@ -148,3 +148,15 @@ items:
       summary: Track plain Name= attribute as well as x:Name for has_named_control triples
       resolution: '2026-04-19: Updated parse_xaml_file() section 3 to collect both x:Name and plain Name= per element into a set before emitting has_named_control triples; added 2 regression tests (plain Name= and same-value dedup). All 189 tests pass.'
       archived_date: "2026-04-19"
+    - key: MINE-KOTLIN-GENERIC-RECEIVER-NESTED
+      summary: Extract function names from Kotlin generic functions with deeply nested receiver types
+      resolution: '2026-04-19: Replaced [^>]+ with (?:[^<>]|<[^<>]*>)* in fun regex; added 2 tests for depth-2 generic nesting (AC-1, AC-2); all 176 symbol extract tests pass'
+      archived_date: "2026-04-19"
+    - key: MINE-BIN-SKIP-DIRS
+      summary: bin/ in SKIP_DIRS silently excludes non-.NET script directories
+      resolution: '2026-04-19: Remove bin from global SKIP_DIRS; add _is_dotnet_project() helper; skip bin/ conditionally in scan_project() when .NET markers present. Hardened round-1: fixed weak .dll test fixture in test_bin_dir_skipped_when_sln_at_root.'
+      archived_date: "2026-04-19"
+    - key: CLARIFY-MINE-BIN-SKIP-DIRS
+      summary: Clarify scope for MINE-BIN-SKIP-DIRS before planning can proceed
+      resolution: '2026-04-19: Owner decision: implement per-project detection — skip bin/ only when .NET markers (.csproj, .sln, .fsproj, .vbproj) are present. Add non-.NET regression test.'
+      archived_date: "2026-04-19"

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/docs/BACKLOG.yaml

@@ -840,26 +840,6 @@ items:
       - [ ] extract_symbol("String getName() {\n}\n", "java") returns ("getName", "method")
       - [ ] extract_symbol("String name;\n", "java") still returns ("", "")
       - [ ] All existing Java symbol-extraction tests pass
-  - key: MINE-KOTLIN-GENERIC-RECEIVER-NESTED
-    summary: Extract function names from Kotlin generic functions with deeply nested receiver types
-    type: feature
-    status: open
-    priority: low
-    size: S
-    section_id: immediate
-    labels: [feat, miner, kotlin]
-    description: |-
-      ## Problem
-      The fun extraction regex uses negated-class [^>] that stops at the first >, so nested generics like Map<String, List<Int>> or bounds like <T : Comparable<T>> cause extract_symbol to return ("", ""). Chunks are stored but symbol metadata is empty, degrading search relevance for generic utility/extension code.
-
-      ## Scope
-      - Update fun regex in _KOTLIN_EXTRACT to handle depth-2 nesting
-      - Cover: type param bounds, generic receivers
-
-      ## Acceptance criteria
-      - [ ] extract_symbol('fun <T : Comparable<T>> List<T>.sorted(): List<T>', 'kotlin') == ('sorted', 'function')
-      - [ ] extract_symbol('fun Map<String, List<Int>>.flatten(): List<Int>', 'kotlin') == ('flatten', 'function')
-      - [ ] All 23 existing Kotlin symbol tests still pass
   - key: MINE-CSHARP-EXPR-BODY
     summary: C# expression-bodied properties not detected as boundaries or symbols
     type: feature
@@ -888,27 +868,6 @@ items:
       - [ ] extract_symbol('public int Count => _items.Count;\n', 'csharp') returns ('Count', 'property')
       - [ ] chunk_code() creates a boundary at expression-bodied property declarations
       - [ ] Existing tests unaffected
-  - key: MINE-BIN-SKIP-DIRS
-    summary: bin/ in SKIP_DIRS silently excludes non-.NET script directories
-    type: task
-    status: open
-    priority: low
-    size: S
-    section_id: immediate
-    labels: [miner]
-    description: |-
-      ## Problem
-      The MINE-DOTNET feature added 'bin' to SKIP_DIRS globally. In non-.NET projects (e.g. Ruby on Rails, Go), a 'bin/' directory legitimately contains source scripts/executables that should be mined. Currently these files are silently skipped with no warning.
-
-      ## Scope
-      - Investigate whether 'bin' should be kept global or made .NET-specific
-      - Option A: Replace global 'bin' entry with per-project detection (only skip bin/ when .csproj/.sln is present in the project root)
-      - Option B: Document as a known limitation in CLAUDE.md and README
-
-      ## Acceptance criteria
-      - A Ruby on Rails or Go project with a 'bin/' directory has those files mined correctly (option A), OR the limitation is clearly documented in user-facing docs (option B)
-      - .NET project mining still skips bin/ and obj/ correctly
-      - No existing tests regress
   - key: BENCH-DOTNET-CLONE-REPO
     summary: Clone jasontaylordev/CleanArchitecture (pin to a stable tag, e.g. v8.0.x) and record the commit hash before running the benchmark
     type: task

{mempalace_code-1.3.0 → mempalace_code-1.4.1}/hooks/README.md

@@ -1,6 +1,6 @@
-# mempalace-code Hooks — Auto-Save for Terminal AI Tools
+# mempalace-code Hooks — Claude Code Auto-Save (Legacy)
 
-These hook scripts make mempalace-code save automatically. No manual "save" commands needed.
+Optional Claude Code hooks that trigger automatic memory saves during conversations. Not required — MCP tools + usage rules (see `docs/AGENT_INSTALL.md` Section 7) achieve the same result for any agent.
 
 ## What They Do
 
@@ -42,24 +42,19 @@ Make them executable:
 chmod +x hooks/mempal_save_hook.sh hooks/mempal_precompact_hook.sh
 ```
 
-## Install — Codex CLI (OpenAI)
+## Codex CLI / Other Agents
 
-Add to `.codex/hooks.json`:
+These hooks are **Claude Code-only** — they rely on Claude Code's `Stop` and `PreCompact` hook events (JSON with `session_id`, `stop_hook_active`, and `transcript_path` on stdin). Other agents (Codex CLI, Cursor, etc.) do not support these events.
 
-```json
-{
-  "Stop": [{
-    "type": "command",
-    "command": "/absolute/path/to/hooks/mempal_save_hook.sh",
-    "timeout": 30
-  }],
-  "PreCompact": [{
-    "type": "command",
-    "command": "/absolute/path/to/hooks/mempal_precompact_hook.sh",
-    "timeout": 30
-  }]
-}
-```
+**What to use instead:**
+
+| Concern | Solution |
+|---------|----------|
+| **Code mining** (indexing source files) | `mempalace watch-all` — works with any client, re-mines on commit |
+| **Conversation context** (decisions, discussions) | MCP tools — `mempalace_add_drawer`, `mempalace_diary_write` |
+| **Session continuity** | Add mempalace usage rules to agent instructions (see `docs/AGENT_INSTALL.md` Section 7) |
+
+For Codex specifically, wire the MCP server in `~/.codex/config.toml` (see `docs/AGENT_INSTALL.md` Step 5.2) and add a system prompt instruction like: *"At the end of each session, save key decisions and context to mempalace using `mempalace_add_drawer` and `mempalace_diary_write`."*
 
 ## Configuration