superlocalmemory 4.1.2 → 4.1.4

package/.claude-plugin/marketplace.json

@@ -21,7 +21,25 @@
       "license": "AGPL-3.0-or-later",
       "name": "superlocalmemory",
       "source": "./plugin",
-      "version": "4.1.2"
+      "version": "4.1.4"
+    },
+    {
+      "author": {
+        "name": "Qualixar"
+      },
+      "description": "Local-first agent memory with auditable hybrid retrieval — the Codex build, with Codex-shaped rules, hooks and launcher.",
+      "homepage": "https://github.com/qualixar/superlocalmemory",
+      "keywords": [
+        "memory",
+        "mcp",
+        "agents",
+        "local-first",
+        "context-compression"
+      ],
+      "license": "AGPL-3.0-or-later",
+      "name": "superlocalmemory-codex",
+      "source": "./codex-plugin",
+      "version": "4.1.4"
     }
   ]
 }

package/CHANGELOG.md

@@ -5,6 +5,71 @@ All notable changes to SuperLocalMemory will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [4.1.4] — A busy daemon is not a dead daemon
+
+### Fixed
+- **`slm serve stop` and `slm restart` could report a live daemon as "not
+  running."** `daemon_request()` preflights every call with `GET /health`,
+  which needs the daemon's single-threaded event loop to be free to answer.
+  `/maintenance/run` and `/consolidate/cognitive` run multi-second (sometimes
+  multi-minute) work directly inline in their handlers with no thread
+  offload — unlike `/recall`, fixed for exactly this in v3.4.52 — so every
+  request blocks for as long as they run, health included. Reproduced live: a
+  genuine `/maintenance/run` call held the loop long enough that 15 of 15
+  health polls in that window timed out at the 2s cap while the process never
+  stopped listening. `stop_daemon()` now checks process liveness directly
+  (PID plus a clock-independent start token, no HTTP) before deciding whether
+  a stop is owed, and sends the stop itself without the health preflight —
+  the daemon still authenticates the request by its capability header on
+  arrival, so no ownership guarantee is lost, only the stall-prone round
+  trip. `cmd_restart`'s Step 1 uses the same liveness check, so a busy-but-alive
+  daemon is no longer skipped as "already stopped," which previously cascaded
+  into Step 3 refusing to start a second daemon and failing the whole restart.
+  The underlying blocking handlers are unchanged by this fix — recall and
+  remember calls can still be delayed by a running maintenance pass, just not
+  misreported as a dead daemon.
+
+### Added
+- **A one-command Antigravity plugin install.** `agy`'s plugin installer only
+  accepts a local directory — there is no marketplace or remote-install
+  path — so Claude Code and Codex had a real one-liner and Antigravity did
+  not. `scripts/install-antigravity-plugin.sh` closes the gap: a shallow,
+  sparse checkout of `antigravity-plugin/` into a local cache, then
+  `agy plugin install` against it. Idempotent; safe to re-run to update.
+
+## [4.1.3] — Your install, your tools, your memories
+
+### Fixed
+- **Installing the plugin could point your editor at a different, empty set of
+  memories.** Its server configuration pinned a data directory of its own, so on
+  a machine already using SLM the editor read somewhere else entirely — measured
+  here, that directory held 28 KB while the real one held 611 MB and 5,370
+  memories. Nothing failed and nothing warned; it simply looked like everything
+  was gone. The plugin no longer says where your memories live.
+- **Installing the plugin could take away tools you had turned on.** Three of the
+  four editor builds forced a narrower tool set — 31 tools, without the eight for
+  coordinating between sessions. If you had deliberately enabled more, installing
+  something removed them. No build states an opinion about this any more.
+- **Installing the plugin could give you a second copy of SuperLocalMemory.** Two
+  launchers preferred a private environment of their own over the one you already
+  had from `pip`, `pipx` or `npm`. So a machine that had ever created one kept
+  using it — including after you upgraded properly — and two copies read the same
+  memories, with whichever started first deciding which one answered. Both now
+  use what you already have, in either install order, and fall back to their own
+  only on a machine with none. When a leftover copy is found beside a real
+  install, it is named, with both version numbers, instead of being used quietly.
+- **The plugin would not start at all outside Claude Code.** The path it built
+  came from a setting only Claude Code provides, so on Codex and Antigravity it
+  pointed nowhere and the server never came up — with no error explaining why.
+- **Codex reported itself as Claude Code.** The catalogue offered one build, the
+  Claude Code one, so Codex installed that and every memory written from Codex was
+  filed under the wrong tool. Codex now has its own entry and its own build,
+  with the Codex-shaped rules, hooks and launcher that were never being delivered.
+
+### Changed
+- **Antigravity is supported.** It had no plugin at all. It now has the same
+  twelve skills, four sub-agents, command and hooks as every other editor.
+
 ## [4.1.2] — The monitor that was watching nothing
 
 ### Fixed

package/README.md

@@ -5,7 +5,7 @@
   </picture>
 </p>
 
-<h1 align="center">SuperLocalMemory V4.1.2</h1>
+<h1 align="center">SuperLocalMemory V4.1.4</h1>
 
 <h2 align="center">Rent the LLM. Own the memory.</h2>
 
@@ -27,12 +27,12 @@ guarantee here is stated as a falsifiable invariant, tested under an adversarial
 negative control, and shipped with the harness that regenerates the evidence:
 <code>python benchmark/run_all.py --trials 200 --output-dir results/</code>. What each experiment
 does <em>not</em> exercise is stated too.</p>
-<p align="center"><code>v4.1.2</code> — one control plane: <strong>SLM-Mesh</strong> peer coordination · multi-scope memory (personal / shared / global) · profiles · Cache · Compress · 7-layer retrieval · code graph · Entity Explorer · skill evolution · Modes A/B/C · GDPR retention &amp; audit chain · bounded loops — across CLI, MCP, dashboard, the <strong>Claude plugin</strong>, the <strong>Codex add-on</strong>, and documented IDE integrations.<br/>
+<p align="center"><code>v4.1.4</code> — one control plane: <strong>SLM-Mesh</strong> peer coordination · multi-scope memory (personal / shared / global) · profiles · Cache · Compress · 7-layer retrieval · code graph · Entity Explorer · skill evolution · Modes A/B/C · GDPR retention &amp; audit chain · bounded loops — across CLI, MCP, dashboard, the <strong>Claude plugin</strong>, the <strong>Codex add-on</strong>, and documented IDE integrations.<br/>
 Proxy: <code>slm wrap claude</code> &nbsp;·&nbsp; MCP: add <code>slm_compress</code> to your config &nbsp;·&nbsp; Skill: zero-config</p>
 <p align="center"><strong>Four public arXiv preprints</strong> · V4: <a href="https://arxiv.org/abs/2608.08253">arXiv:2608.08253</a> · companion archive: <a href="https://zenodo.org/records/21853302">Zenodo 21853302</a> (<a href="https://doi.org/10.5281/zenodo.21853302">DOI 10.5281/zenodo.21853302</a>) · prior preprints: <a href="https://arxiv.org/abs/2603.02240">2603.02240</a> · <a href="https://arxiv.org/abs/2603.14588">2603.14588</a> · <a href="https://arxiv.org/abs/2604.04514">2604.04514</a>.</p>
 
 <p align="center">
-  <a href="CHANGELOG.md"><img src="https://img.shields.io/badge/v4.1.2-Current_Release-2ea44f?style=for-the-badge&logo=checkmarx&logoColor=white" alt="v4.1.2 — Current Release"/></a>
+  <a href="CHANGELOG.md"><img src="https://img.shields.io/badge/v4.1.4-Current_Release-2ea44f?style=for-the-badge&logo=checkmarx&logoColor=white" alt="v4.1.4 — Current Release"/></a>
   <a href="https://arxiv.org/abs/2608.08253"><img src="https://img.shields.io/badge/arXiv-2608.08253-b31b1b?style=for-the-badge&logo=arxiv&logoColor=white" alt="SuperLocalMemory 4.0 paper on arXiv:2608.08253"/></a>
   <a href="https://zenodo.org/records/21853302"><img src="https://img.shields.io/badge/Zenodo-10.5281%2Fzenodo.21853302-1682D4?style=for-the-badge&logo=zenodo&logoColor=white" alt="V4 paper on Zenodo: 10.5281/zenodo.21853302"/></a>
   <a href="https://arxiv.org/abs/2603.14588"><img src="https://img.shields.io/badge/arXiv-2603.14588-b31b1b?style=for-the-badge&logo=arxiv&logoColor=white" alt="arXiv Paper"/></a>

package/ide/configs/codex-mcp.toml

@@ -8,4 +8,6 @@
 [mcp_servers.superlocalmemory]
 command = "slm"
 args = ["mcp"]
-env = { SLM_MCP_PROFILE = "code", SLM_AGENT_ID = "codex" }
+# Agent id only. Pinning a profile here would remove tools you had enabled;
+# unset resolves to the full tool set.
+env = { SLM_AGENT_ID = "codex" }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "superlocalmemory",
-  "version": "4.1.2",
+  "version": "4.1.4",
   "description": "Local-first agent memory with MCP and an agent-native CLI. Documented clients include Claude Code, Cursor, and Windsurf.",
   "keywords": [
     "ai-memory",

package/plugin/.claude-plugin/plugin.json

@@ -15,5 +15,5 @@
   "mcpServers": "./.mcp.json",
   "name": "superlocalmemory",
   "repository": "https://github.com/qualixar/superlocalmemory",
-  "version": "4.1.2"
+  "version": "4.1.4"
 }

package/plugin/.mcp.json

@@ -4,9 +4,7 @@
       "command": "${CLAUDE_PLUGIN_ROOT}/scripts/slm-launch",
       "args": [],
       "env": {
-        "SLM_MCP_PROFILE": "code",
-        "SLM_AGENT_ID": "claude_code",
-        "SLM_DATA_DIR": "${CLAUDE_PLUGIN_DATA}"
+        "SLM_AGENT_ID": "claude_code"
       }
     }
   }

package/plugin/CLAUDE.md

@@ -1,4 +1,4 @@
-<!-- BEGIN SuperLocalMemory v4.1.2 -->
+<!-- BEGIN SuperLocalMemory v4.1.4 -->
 
 ## SuperLocalMemory (SLM) — Agent Rules
 
@@ -39,6 +39,6 @@ slm-recall · slm-remember · slm-session · slm-status · slm-cache · slm-comp
 ### Subagents
 slm-memory-advisor (memory decisions, session hygiene, scope/profile guidance) · slm-optimize-advisor (context compression + KV cache) · slm-governance-advisor (scope/roles/compliance/GDPR)
 
-<!-- END SuperLocalMemory v4.1.2 -->
+<!-- END SuperLocalMemory v4.1.4 -->
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/agents/slm-governance-advisor.md

@@ -77,4 +77,4 @@ slm-scope · slm-governance · slm-profile · slm-remember · slm-recall
 # What NOT to do
 Never session_init twice; never forget without dry-run preview; never store secrets; never bypass role checks; never claim an erasure succeeded without verifying via recall.
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/agents/slm-loop-runner.md

@@ -68,4 +68,4 @@ assessment. The gate is the authority.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/agents/slm-memory-advisor.md

@@ -46,4 +46,4 @@ slm-recall · slm-remember · slm-session · slm-scope · slm-profile · slm-gov
 # What NOT to do
 Never session_init twice; never forget dry_run=False without reporting preview; never dump a whole file into remember; never invent a memory; never claim "saved" without success:true / clean CLI exit; never bypass scope or governance restrictions.
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/agents/slm-optimize-advisor.md

@@ -41,4 +41,4 @@ slm-compress · slm-cache · slm-status · slm-profile
 # What NOT to do
 Never compress code-for-edit/JSON-to-parse/<500 chars; never store secrets/ccr_ids; never let optimize failure block/alter the task; never claim a specific savings %; never carry ccr_ids across profile switches.
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/requirements.txt

@@ -1 +1 @@
-superlocalmemory==4.1.2
+superlocalmemory==4.1.4

package/plugin/scripts/slm-launch

@@ -1,62 +1,131 @@
 #!/usr/bin/env bash
-# slm-launch — WP-F SuperLocalMemory MCP launcher (POSIX)
+# slm-launch — SuperLocalMemory MCP launcher (POSIX)
 #
 # Cross-platform counterpart: slm-launch.bat (Windows)
-# Referenced by plugin/.mcp.json as the MCP server command.
+# Referenced by .mcp.json as the MCP server command.
 #
-# Resolves the correct slm binary for POSIX and joins the namespace daemon
-# before opening the MCP stdio transport. This preserves one writer for all
-# parallel Claude sessions that share CLAUDE_PLUGIN_DATA.
+# WHAT THIS RESOLVES, AND WHY THE DEFAULT CHANGED
 #
-# On Windows, Claude Code invokes slm-launch.bat instead (same-stem, .bat extension).
+# It used to default to a venv the plugin owns:
+# "${CLAUDE_PLUGIN_DATA}/venv/bin/slm". So anyone who already had SuperLocalMemory
+# — from pip, pipx or npm — got a SECOND installation the moment they installed
+# the plugin, on its own schedule, upgraded separately, and easy to leave behind
+# a release. Installing a plugin should not fork your installation.
 #
-# Environment:
-#   CLAUDE_PLUGIN_DATA — persistent data dir where the plugin venv lives
-#   SLM_DATA_DIR       — optional; SLM data dir (default: ~/.superlocalmemory)
-#   SLM_LAUNCHER       — optional; which slm binary to use:
-#                          plugin (default) — the plugin venv binary
-#                          system           — the system `slm` on PATH
-#                          <path>           — an explicit slm binary (leading ~ expanded)
+# It was also Claude-Code-only. ${CLAUDE_PLUGIN_DATA} is set by Claude Code and
+# by nothing else, so on Codex or Antigravity the path collapsed to
+# "/venv/bin/slm" and the server simply did not start.
+#
+# The default is now: USE THE SLM THAT IS ALREADY INSTALLED. A plugin-owned venv
+# is the fallback for a machine that has none, not the first choice.
+#
+# Environment (all optional):
+#   SLM_LAUNCHER       auto (default) — prefer an installed slm, else the plugin venv
+#                      system         — require the slm on PATH
+#                      plugin         — require the plugin's own venv
+#                      <path>         — an explicit slm binary (leading ~ expanded)
+#   SLM_DATA_DIR       SLM data dir (default: ~/.superlocalmemory)
+#   CLAUDE_PLUGIN_DATA where a plugin-owned venv lives, when there is one
+#
+# Deliberately NOT set here: the MCP profile. Which tools a host exposes is the
+# user's decision, and a plugin that narrows it takes away tools they configured.
 
 _SLM_DATA="${SLM_DATA_DIR:-$HOME/.superlocalmemory}"
+_PLUGIN_BIN="${CLAUDE_PLUGIN_DATA:+${CLAUDE_PLUGIN_DATA}/venv/bin/slm}"
 
-# If a system daemon is already running (its pidfile points at a live process),
-# there is no need to boot the plugin's own venv daemon — we can open an MCP
-# writer against the daemon that is already up. This skips a redundant, slow
-# venv bootstrap on every session start.
+# Is a daemon already up for this data dir? If so we join it rather than boot a
+# second one — one writer for every session that shares the store.
 _daemon_running=0
 if [ -f "${_SLM_DATA}/daemon.pid" ] \
    && kill -0 "$(cat "${_SLM_DATA}/daemon.pid" 2>/dev/null)" 2>/dev/null; then
     _daemon_running=1
 fi
 
-case "${SLM_LAUNCHER:-plugin}" in
-    plugin)
-        SLM_BIN="${CLAUDE_PLUGIN_DATA}/venv/bin/slm"
-        # Only bootstrap the owned daemon when one is not already running.
-        if [ "${_daemon_running}" -eq 0 ]; then
-            if ! "${SLM_BIN}" serve start >&2; then
+# Expand a leading ~ WITHOUT eval — eval on an environment value is a
+# command-injection foot-gun. Only tilde expansion; pass expanded paths for vars.
+_expand_tilde() {
+    case "$1" in
+        "~"*) printf '%s' "${HOME}${1#\~}" ;;
+        *)    printf '%s' "$1" ;;
+    esac
+}
+
+# Start a daemon with the chosen binary unless one is already serving.
+_ensure_daemon() {
+    [ "${_daemon_running}" -eq 1 ] && return 0
+    "$1" serve start >&2
+}
+
+case "${SLM_LAUNCHER:-auto}" in
+    auto)
+        # 1. An slm already on PATH — the pip, pipx or npm install the user has.
+        #    This is the common case and the one that must not be forked.
+        if command -v slm >/dev/null 2>&1; then
+            # Both orderings end up here, and both must converge on ONE
+            # environment:
+            #   installed first, plugin second -> nothing was ever forked;
+            #   plugin first, installed second -> the plugin venv is now the
+            #     older copy and must stop being used.
+            # They already share one store, because this launcher sets no
+            # SLM_DATA_DIR and both resolve ~/.superlocalmemory. What is left to
+            # go wrong is two DIFFERENT VERSIONS serving that one store: whoever
+            # starts the daemon first decides which code the other one talks to.
+            # So when a leftover plugin venv is found beside a real install, say
+            # so, with both versions named. Silent duplication is how a store
+            # ends up served by a release nobody thinks they are running.
+            if [ -n "${_PLUGIN_BIN}" ] && [ -x "${_PLUGIN_BIN}" ]; then
+                _sys_v="$(slm --version 2>/dev/null | awk '{print $NF}')"
+                _plg_v="$("${_PLUGIN_BIN}" --version 2>/dev/null | awk '{print $NF}')"
+                if [ -n "${_plg_v}" ] && [ "${_sys_v}" != "${_plg_v}" ]; then
+                    echo "SLM plugin: using the installed slm ${_sys_v:-?} and IGNORING a leftover plugin venv at ${_PLUGIN_BIN} (${_plg_v})." >&2
+                    echo "  One store, one version. Remove the venv to stop it drifting:  rm -rf \"${CLAUDE_PLUGIN_DATA}/venv\"" >&2
+                fi
+            fi
+            _ensure_daemon "$(command -v slm)" || \
+                echo "SLM plugin: daemon did not start; continuing as a direct MCP writer." >&2
+            exec slm mcp
+        fi
+        # 2. No system install. Use the plugin's own venv if this host gave us
+        #    somewhere to keep one.
+        if [ -n "${_PLUGIN_BIN}" ] && [ -x "${_PLUGIN_BIN}" ]; then
+            if ! _ensure_daemon "${_PLUGIN_BIN}"; then
                 echo "SLM plugin: unable to start the owned daemon; refusing a direct MCP writer." >&2
                 exit 1
             fi
+            exec "${_PLUGIN_BIN}" mcp
         fi
-        exec "${SLM_BIN}" mcp
+        # 3. Neither. Say which two things were looked for, so the fix is obvious.
+        echo "SLM plugin: no SuperLocalMemory found." >&2
+        echo "  Looked for: 'slm' on PATH, and ${_PLUGIN_BIN:-a plugin venv (this host sets no plugin data dir)}." >&2
+        echo "  Install it with:  pipx install superlocalmemory" >&2
+        exit 1
         ;;
     system)
+        if ! command -v slm >/dev/null 2>&1; then
+            echo "SLM_LAUNCHER=system but no 'slm' on PATH." >&2
+            exit 1
+        fi
+        _ensure_daemon "$(command -v slm)" || true
         exec slm mcp
         ;;
+    plugin)
+        if [ -z "${_PLUGIN_BIN}" ]; then
+            echo "SLM_LAUNCHER=plugin but this host sets no CLAUDE_PLUGIN_DATA." >&2
+            exit 1
+        fi
+        if ! _ensure_daemon "${_PLUGIN_BIN}"; then
+            echo "SLM plugin: unable to start the owned daemon; refusing a direct MCP writer." >&2
+            exit 1
+        fi
+        exec "${_PLUGIN_BIN}" mcp
+        ;;
     *)
-        # Explicit binary path. Expand a leading ~ to $HOME WITHOUT eval — eval
-        # on an environment value is a command-injection foot-gun. Only tilde
-        # expansion is supported here; pass an already-expanded path for $VARs.
-        case "${SLM_LAUNCHER}" in
-            "~"*) _SLM_BIN="${HOME}${SLM_LAUNCHER#\~}" ;;
-            *)    _SLM_BIN="${SLM_LAUNCHER}" ;;
-        esac
+        _SLM_BIN="$(_expand_tilde "${SLM_LAUNCHER}")"
         if [ ! -x "${_SLM_BIN}" ]; then
             echo "SLM_LAUNCHER is not an executable slm binary: ${_SLM_BIN}" >&2
             exit 1
         fi
+        _ensure_daemon "${_SLM_BIN}" || true
         exec "${_SLM_BIN}" mcp
         ;;
 esac

package/plugin/skills/slm-cache/SKILL.md

@@ -145,4 +145,4 @@ These subcommands control daemon-level cache settings. They do not read or write
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/skills/slm-compress/SKILL.md

@@ -147,4 +147,4 @@ Content over 1 MB (1 000 000 bytes UTF-8) is processed but `reversible` is force
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/skills/slm-governance/SKILL.md

@@ -245,4 +245,4 @@ Before running any destructive operation (`forget`, `compact_memories`):
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-graph/SKILL.md

@@ -312,4 +312,4 @@ profile. See `slm-profile` for the full profile switching workflow.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/skills/slm-loop/SKILL.md

@@ -96,4 +96,4 @@ paused, name the approval needed; when errored, quote the short detail.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin/skills/slm-mesh/SKILL.md

@@ -279,4 +279,4 @@ mesh availability.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-profile/SKILL.md

@@ -146,4 +146,4 @@ Name them differently in your MCP config (e.g. `superlocalmemory-personal` and
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-recall/SKILL.md

@@ -323,4 +323,4 @@ before recalling, then switch back. See `slm-profile` for workspace switching.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-remember/SKILL.md

@@ -270,4 +270,4 @@ different workspace, use `switch_profile` first. See `slm-profile`.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-scope/SKILL.md

@@ -173,4 +173,4 @@ to review the impact. See `slm-remember` for the full deletion discipline.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-session/SKILL.md

@@ -253,4 +253,4 @@ explicitly and call `recall` with `include_global`/`include_shared` after
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin/skills/slm-status/SKILL.md

@@ -163,4 +163,4 @@ multi-profile setup. To switch the active profile, see `slm-profile`.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/rules/AGENTS.md

@@ -137,4 +137,4 @@ When the SLM MCP server is unavailable, use these CLI equivalents:
 - **slm-optimize-advisor** — context compression and KV cache
 - **slm-governance-advisor** — scope/role compliance, retention policies, GDPR
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/skills/slm-cache/SKILL.md

@@ -145,4 +145,4 @@ These subcommands control daemon-level cache settings. They do not read or write
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/skills/slm-compress/SKILL.md

@@ -147,4 +147,4 @@ Content over 1 MB (1 000 000 bytes UTF-8) is processed but `reversible` is force
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/skills/slm-governance/SKILL.md

@@ -245,4 +245,4 @@ Before running any destructive operation (`forget`, `compact_memories`):
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-graph/SKILL.md

@@ -312,4 +312,4 @@ profile. See `slm-profile` for the full profile switching workflow.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/skills/slm-loop/SKILL.md

@@ -96,4 +96,4 @@ paused, name the approval needed; when errored, quote the short detail.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/plugin-src/skills/slm-mesh/SKILL.md

@@ -279,4 +279,4 @@ mesh availability.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-profile/SKILL.md

@@ -146,4 +146,4 @@ Name them differently in your MCP config (e.g. `superlocalmemory-personal` and
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-recall/SKILL.md

@@ -323,4 +323,4 @@ before recalling, then switch back. See `slm-profile` for workspace switching.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-remember/SKILL.md

@@ -270,4 +270,4 @@ different workspace, use `switch_profile` first. See `slm-profile`.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-scope/SKILL.md

@@ -173,4 +173,4 @@ to review the impact. See `slm-remember` for the full deletion discipline.
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-session/SKILL.md

@@ -253,4 +253,4 @@ explicitly and call `recall` with `include_global`/`include_shared` after
 
 ---
 
-*SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later*

package/plugin-src/skills/slm-status/SKILL.md

@@ -163,4 +163,4 @@ multi-profile setup. To switch the active profile, see `slm-profile`.
 
 ---
 
-SuperLocalMemory v4.1.2 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.1.4 · Qualixar · AGPL-3.0-or-later

package/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "superlocalmemory"
-version = "4.1.2"
+version = "4.1.4"
 description = "Local-first agent memory with auditable hybrid retrieval"
 readme = "README.md"
 license = "AGPL-3.0-or-later"

package/src/superlocalmemory/__init__.py

@@ -32,7 +32,7 @@ if "OMP_NUM_THREADS" not in os.environ:
     os.environ["OMP_NUM_THREADS"] = "2"
 # ---------------------------------------------------------------------------
 
-__version__ = "4.1.2"
+__version__ = "4.1.4"
 
 _REQUIRED_VERSIONS = {
     "sentence_transformers": "5.3.0",

package/src/superlocalmemory/cli/commands.py

@@ -754,13 +754,19 @@ def cmd_restart(args: Namespace) -> None:
     # Step 1: stop only the descriptor-owned daemon. Its graceful shutdown
     # owns worker termination; process-name-wide scans are forbidden.
     from superlocalmemory.cli.daemon import (
-        is_daemon_running,
+        owned_daemon_process_alive,
         read_descriptor,
         stop_daemon,
         wait_for_owned_daemon_shutdown,
     )
 
-    was_running = is_daemon_running()
+    # v4.1.4: process liveness, not HTTP health-readiness, decides whether a
+    # stop is owed. is_daemon_running() additionally requires a live /health
+    # response, which a daemon busy inside /maintenance/run or
+    # /consolidate/cognitive cannot give for the duration of that call even
+    # though it is fully alive and holding the port — see stop_daemon()'s
+    # docstring for the reproduced failure this fixes.
+    was_running = owned_daemon_process_alive()
     owned_descriptor = read_descriptor() if was_running else None
     stopped = stop_daemon() if was_running else True
     if stopped and was_running:

package/src/superlocalmemory/cli/daemon.py

@@ -254,6 +254,29 @@ def is_daemon_running() -> bool:
     return legacy is not None
 
 
+def owned_daemon_process_alive() -> bool:
+    """Return whether an owned daemon *process* is alive, HTTP aside.
+
+    ``is_daemon_running()`` additionally requires a live, matching
+    ``/health`` response, which conflates two different questions: "is there
+    a process I need to stop" and "is it ready to serve requests right now."
+    A daemon whose event loop is synchronously blocked by a long-running
+    handler (``/maintenance/run``, ``/consolidate/cognitive``) cannot answer
+    the second for the duration of that call, but the answer to the first is
+    still yes. Callers that only need to decide whether a stop is owed
+    (``slm restart`` Step 1) should use this instead, so a transiently busy
+    daemon is not skipped as "already stopped" while it keeps running and
+    holding the port — which then made Step 3 refuse to start a second
+    daemon on the still-occupied port and fail the whole restart.
+    """
+    descriptor = read_descriptor()
+    if descriptor is not None:
+        return _descriptor_process_is_alive(descriptor)
+    if descriptor_path().exists():
+        return False
+    return _verified_legacy_health() is not None
+
+
 def _fetch_health(port: int) -> dict | None:
     """Fetch loopback health without following cross-namespace discovery."""
     try:
@@ -349,8 +372,25 @@ def daemon_request(
     timeout_seconds: float = 30.0,
     expected_descriptor=_EXPECTED_DESCRIPTOR_UNSET,
     expected_legacy: dict | None = None,
+    verify_health: bool = True,
 ) -> dict | None:
-    """Send a request only after validating the owned daemon identity."""
+    """Send a request only after validating the owned daemon identity.
+
+    ``verify_health`` — when True (the default), a ``GET /health`` preflight
+    must succeed and match the descriptor before the real request is sent.
+    That preflight needs the daemon's event loop to be free to answer HTTP,
+    which is a *readiness* question, not a *liveness* one: a daemon whose
+    loop is synchronously blocked by a long-running handler (e.g.
+    ``/maintenance/run``, ``/consolidate/cognitive``, neither of which is
+    offloaded to a thread the way ``/recall`` was for exactly this reason in
+    v3.4.52) cannot answer /health for the duration of that call even though
+    the process is fully alive and listening. Callers that have already
+    proven process-level ownership some other way (e.g. ``stop_daemon()`` via
+    ``_descriptor_process_is_alive``) should pass ``verify_health=False`` so a
+    busy-but-alive daemon does not get misreported as not running. Only
+    meaningful for the descriptor path — the legacy bridge has no capability
+    header and still needs health to identify its target.
+    """
     legacy = None
     if expected_legacy is not None:
         # Legacy daemons have no capability header. Bind the compatibility
@@ -377,11 +417,12 @@ def daemon_request(
     capability: str | None = None
     target_instance: str | None = None
     if descriptor is not None:
-        health = _fetch_health(descriptor.port)
-        if health is None or not descriptor_matches_health(descriptor, health):
-            return None
-        if method.upper() == "GET" and path == "/health":
-            return health
+        if verify_health:
+            health = _fetch_health(descriptor.port)
+            if health is None or not descriptor_matches_health(descriptor, health):
+                return None
+            if method.upper() == "GET" and path == "/health":
+                return health
         port = descriptor.port
         capability = descriptor.capability
         target_instance = descriptor.instance_id
@@ -796,16 +837,38 @@ def stop_daemon() -> bool:
     HTTP capability; the daemon itself terminates its child process tree.
     Success means the owned process exited and released its listener, not just
     that the asynchronous stop request was accepted.
+
+    A busy daemon is not a dead daemon. ``daemon_request()`` normally
+    preflights every call with ``GET /health`` before sending it, but that
+    preflight needs the daemon's single-threaded event loop to be free to
+    answer HTTP. ``/maintenance/run`` and ``/consolidate/cognitive`` run
+    multi-second (sometimes multi-minute) synchronous work directly inline in
+    their handlers with no thread offload, which blocks *every* request on
+    that loop, health included, for as long as they run. Reproduced live: a
+    genuine ``/maintenance/run`` call held the loop long enough that 15/15
+    health polls during the window timed out at exactly the 2s cap while
+    ``ps``/``lsof`` proved the process never stopped listening — which is
+    exactly the "Daemon was not running" false report this fixes. Process
+    liveness (PID + clock-independent start token, proven below via
+    ``_descriptor_process_is_alive``) is the fact that actually matters for
+    "should I try to stop this," so it is checked directly and the mutating
+    ``/stop`` POST is sent with ``verify_health=False`` once that is proven —
+    the daemon still authenticates the request by its capability header on
+    arrival, so this loses no ownership guarantee, only the redundant,
+    stall-prone preflight round trip.
     """
     descriptor = read_descriptor()
     legacy = _verified_legacy_health() if descriptor is None else None
     if descriptor is None and legacy is None:
         return False
     if descriptor is not None:
+        if not _descriptor_process_is_alive(descriptor):
+            return False
         response = daemon_request(
             "POST",
             "/stop",
             expected_descriptor=descriptor,
+            verify_health=False,
         )
     else:
         if legacy is None: