superlocalmemory 4.2.4 → 4.2.6

package/.claude-plugin/marketplace.json

@@ -9,7 +9,7 @@
       "author": {
         "name": "Qualixar"
       },
-      "description": "Local-first agent memory + reversible context compression and KV cache, as an MCP server, with optional pictures, PDFs and bot messaging. Serves the 58-tool full profile by default; set SLM_MCP_PROFILE=code for the 39-tool code profile.",
+      "description": "Local-first agent memory + reversible context compression and KV cache, as an MCP server, with optional pictures, PDFs and bot messaging. Serves the 58-tool full profile by default; `slm tools set <name>` changes it for every app.",
       "homepage": "https://github.com/qualixar/superlocalmemory",
       "keywords": [
         "memory",
@@ -21,7 +21,7 @@
       "license": "AGPL-3.0-or-later",
       "name": "superlocalmemory",
       "source": "./plugin",
-      "version": "4.2.4"
+      "version": "4.2.6"
     },
     {
       "author": {
@@ -39,7 +39,7 @@
       "license": "AGPL-3.0-or-later",
       "name": "superlocalmemory-codex",
       "source": "./codex-plugin",
-      "version": "4.2.4"
+      "version": "4.2.6"
     }
   ]
 }

package/.cursor-plugin/marketplace.json

@@ -0,0 +1,29 @@
+{
+  "metadata": {
+    "description": "SuperLocalMemory: local-first long-term memory for agents and bots."
+  },
+  "name": "qualixar",
+  "owner": {
+    "email": "varun.pratap.bhardwaj@gmail.com",
+    "name": "Qualixar"
+  },
+  "plugins": [
+    {
+      "category": "productivity",
+      "description": "Local-first long-term memory for your agents and bots: remember, recall and session context stored on your own computer, with reversible compression and a cache for large tool outputs.",
+      "homepage": "https://superlocalmemory.com",
+      "keywords": [
+        "memory",
+        "long-term-memory",
+        "local-first",
+        "mcp",
+        "agents"
+      ],
+      "license": "MIT",
+      "name": "superlocalmemory",
+      "repository": "https://github.com/qualixar/superlocalmemory",
+      "source": "./cursor-plugin",
+      "version": "4.2.6"
+    }
+  ]
+}

package/CHANGELOG.md

@@ -5,6 +5,171 @@ 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).
 
+## [4.2.6] — The Hermes plugin works on a default install, and every host package meets its directory's rules
+
+No migration. Nothing is moved or deleted.
+
+### Hermes
+
+- **The plugin did nothing on most installs.** Hermes refuses a plugin's MCP
+  calls until `config.yaml` grants the server under
+  `plugins.entries.superlocalmemory.mcp_allowlist`, and it does not ask for
+  that grant; the install guide said it would. The plugin treated the refusal
+  like any failure, so recall, session tracking, tool logging and turn capture
+  were silently off, and `slm upgrade-hosts` reported Hermes as verified. Now:
+  - `slm upgrade-hosts --host hermes --apply` writes the grant (a backup is
+    kept; a config with hand-written comments is never rewritten, you are shown
+    the line instead);
+  - `slm doctor` has a "Hermes plugin access" check;
+  - the plugin warns once in Hermes's log with the fix, and `/slm status` leads
+    with it;
+  - the guide shows the step.
+- **`capture_turns: true` kept almost nothing (#263).** Turns went to the
+  auto-capture path, which keeps only English sentences about a decision, a bug
+  or a preference. Every completed turn is now saved as a memory, in any
+  language, with secrets and email addresses redacted first; a retried turn is
+  saved once.
+
+Checked on a real Hermes: its hooks (tool logging, turn capture, recall from an
+earlier session), 16 skills, the `/slm` commands and 4 agent tools.
+
+### Plugin directories
+
+- **Claude Code:** the plugin passes `claude plugin validate --strict`. It now
+  has a README stating what each hook does and where data goes, its licence
+  file, and licence and homepage in the manifest. The agent rules no longer sit
+  in a root `CLAUDE.md` that Claude Code never loaded; the skills teach them.
+- **Cursor and Grok Bot:** the Cursor-format plugin is its own folder,
+  `cursor-plugin/`, under the MIT licence (Cursor's marketplace does not accept
+  copyleft). It is a launcher for the separately installed engine, which stays
+  AGPL-3.0-or-later.
+- **VS Code and Antigravity IDE extension:** a new MIT extension,
+  `vscode-extension/`, adds SuperLocalMemory's MCP server to the editor
+  (VS Code 1.101 or later) when `slm` is installed. It adds nothing when that
+  editor's own MCP config already has SuperLocalMemory (VS Code's `mcp.json`,
+  Antigravity's `mcp_config.json`), so there is never a second copy. Cursor
+  does not support the VS Code MCP provider API the extension uses; in Cursor
+  use the Cursor plugin or `slm connect cursor`. The `.vsix` is attached to the GitHub release. (A first build
+  left one of its two modules out of the package, so it failed to start in
+  VS Code; a test now checks every module it loads is packaged.)
+- **Codex:** the plugin carries its icon, logo, privacy and terms links.
+- **Official MCP Registry:** `server.json` ships in the repository and moves
+  with every release.
+
+### One tool set for every app
+
+- **`slm tools`** shows which MCP tool set your apps get and how many tools
+  each set has; **`slm tools set <name>`** changes it for every app at once
+  (`slm tools set default` goes back). Restart the apps to see it. An app whose
+  own config sets `SLM_MCP_PROFILE` keeps that value.
+- The default stays **`full`, 58 tools**: all of memory, SLM-Mesh, optimize and
+  loops, under the roughly 100-tool limit some hosts put on all MCP servers
+  together. No plugin or extension forces a set any more: the Cursor and Grok
+  Bot plugin used to give 19 tools and 6 skills (now the default and all 16
+  skills), the Antigravity plugin 70.
+- The code-graph and governance skills now tell the agent what to say when
+  their tools are not in the current set (`slm tools set code` or `power`).
+
+### New hosts
+
+- **Grok CLI:** `slm connect grok-cli` writes the MCP server into
+  `~/.grok/config.toml`, and Grok CLI runs the Claude Code plugin as is
+  (`grok plugin install <path>/plugin --trust`): 16 skills, 4 agents and the
+  hooks. One hook did not start there, because Grok runs a command that is only
+  a quoted path as a file name with the quotes kept; every hook command now
+  starts in Grok CLI and Claude Code alike. Grok sends hook data in camelCase
+  (`sessionId`, `toolName`); every `slm hook` now reads both spellings, so tool
+  outcomes are recorded there too. `slm connect grok-cli` honours `$GROK_HOME`.
+- **Muse CLI:** `slm connect muse` writes the MCP server into Muse's
+  `settings.json` and keeps every other setting. A file that still has the old
+  `mcp_servers` key gets it renamed to `mcpServers` in the same write, since
+  Muse ignores every server when both are present. A new `muse-plugin/` brings
+  the 16 skills, 4 agents, `/slm-loop` and the hooks: the Claude Code plugin
+  without the `allowed-tools` lines Muse rejects and without an MCP launcher,
+  and with a session-start hook that answers in JSON (Muse drops output that
+  starts with `[`, which the plain-text one does). Its MCP entry is marked
+  optional, so Muse never aborts a run because SuperLocalMemory did not start,
+  and its agents name the tools the way Muse does (`mcp__superlocalmemory__…`).
+
+- **VS Code's Agents window and GitHub Copilot CLI:** `slm connect copilot-cli`
+  writes `~/.copilot/mcp-config.json`, which both read (the Agents window does
+  not see servers from editor extensions). `vscode-agent-plugin/`, listed by
+  `.github/plugin/marketplace.json` where VS Code looks first, brings the skills
+  and agents with a plain `slm mcp` server: the Claude Code plugin's launcher
+  uses a shell default VS Code cannot run.
+- **OpenClaw:** `slm connect openclaw` adds the server under `mcp.servers` in
+  `openclaw.json` and copies 14 skills into `~/.agents/skills`, where OpenClaw
+  loads them. A JSON5 file with comments is never rewritten; the command to run
+  is printed instead.
+
+### Packaging
+
+- **The npm package would have shipped no Cursor plugin.** It moved to its own
+  `cursor-plugin/` folder in this release, and `package.json` still listed only
+  the old folders. Both `cursor-plugin/` and `.cursor-plugin/` ship now, and a
+  test checks every plugin folder the guide points at.
+
+### Fixes
+
+- **Rolling back the text model could start the picture model's worker**
+  (#262). A previous model saved under the picture worker's provider was loaded
+  through that worker, which already held its own model; on a small machine the
+  rollback stopped for lack of memory. Rollback now loads it under the automatic
+  provider. The CLI and the dashboard share this path.
+
+## [4.2.5] — Every sentence you save can be found, and saves no longer hold up the store
+
+No migration. Nothing is moved or deleted.
+
+### Every sentence is findable
+
+- **A long memory kept only a few of its sentences as facts.** A 767-character
+  launch note was saved as the whole note plus four facts, its first two
+  bullets; asking "Product Hunt license fix" found nothing, because that line
+  lived only inside the whole-note fact. Three hard-coded limits of five facts
+  per memory overrode the configured ten. The configured limit is back, and
+  after extraction every sentence or bullet that no extracted fact already
+  holds is kept as a fact of its own, word for word (up to 40 per memory;
+  greetings, fragments and markup lines are skipped). Word-for-word text cannot
+  misstate what was saved, so it needs no model call.
+- **Memories saved before this release can gain the same**, and this is off
+  by default. On a 27,700-fact store it adds about 14,800 facts (half again
+  the store), and its effect on recall time has not yet been measured on a
+  quiet machine, so it is not switched on for everyone. To turn it on, add
+  `"encoding": {"coverage_backfill_per_memory": 10}` to `config.json` in the
+  data folder and run `slm restart`. It then adds up to ten facts per memory,
+  newest first, in small batches that give way to every recall. It only
+  touches memories whose full text is already searchable, so nothing becomes
+  findable that was not before; anything erased, archived, corrected or held
+  for review is left alone. Each added fact is listed in
+  `backfill/coverage-backfill.jsonl` in the data folder.
+
+### Saves no longer hold up the store
+
+On a copy of a 27,700-fact store, two steps of every save scanned the whole
+store while holding the write lock, so under load other writes waited, mesh
+heartbeats failed with "database is locked" and accepted memories retried:
+
+| Step, same store | 4.2.4 | 4.2.5 |
+|---|---|---|
+| Duplicate check on each save | 1.3-1.7 s (5-30 s under load) | under 1 ms |
+| Keeping a scene's member list (1,607 members) | 4.26 s | 16 ms |
+| Lock holds over 1.5 s while 52 memories were saved (load 25-53) | 41, longest 5.9 s | 0 |
+
+- A write whose caller has already given up now stops at its deadline and lets
+  go of the lock, instead of finishing work that is rolled back anyway.
+- The background re-read of upcoming plans reads and classifies outside the
+  write lock; it locks only for its own update.
+
+### Fixes
+
+- **`slm doctor` warned about memories saved minutes ago** as "missing from the
+  search index". They are indexed in the background, a minute or two behind
+  on a busy machine. Memories from the last ten minutes are now reported as
+  being indexed; older ones still warn.
+- **Repair checks were skipped on a machine up less than half an hour.** "Never
+  checked" was stored as 0 and compared with a clock that counts from boot.
+
 ## [4.2.4] — Recall in under a second on large memories, with the answer check on every question
 
 A performance release. Nothing is moved or deleted. One data-only migration

package/README.md

@@ -1,3 +1,4 @@
+<!-- mcp-name: io.github.qualixar/superlocalmemory -->
 <div align="center">
 
 <picture>
@@ -19,7 +20,7 @@
 [![License: AGPL-3.0](https://img.shields.io/badge/license-AGPL--3.0-blue)](https://github.com/qualixar/superlocalmemory/blob/main/LICENSE)
 [![Research: 4 arXiv papers](https://img.shields.io/badge/arXiv-4_papers-b31b1b)](#research)
 
-**[Install](#install)** · **[See it work](#30-second-example)** · **[Product film](https://www.youtube.com/watch?v=pyAu_vaIYZQ)** · **[Every capability](#everything-slm-does)** · **[Docs](#documentation)** · **[Website](https://www.superlocalmemory.com)**
+**[Install](#install)** · **[See it work](#30-second-example)** · **[Full demo](https://www.youtube.com/watch?v=9Y0wAT07KTc)** · **[Every capability](#everything-slm-does)** · **[Docs](#documentation)** · **[Website](https://www.superlocalmemory.com)**
 
 <img src="https://raw.githubusercontent.com/qualixar/superlocalmemory/main/docs/assets/gifs/recall-tiers.gif" alt="Recall in the dashboard: a real question gets its answer; a gibberish question gets 'That isn't in SuperLocalMemory.'; a real question memory cannot answer exactly gets the closest memories, labelled" width="820">
 
@@ -131,12 +132,40 @@ SuperLocalMemory is part of Qualixar's AI Reliability Engineering work: agent me
 
 <div align="center">
 
-[![Watch the SuperLocalMemory product film: Claude Code and Codex sharing one memory](https://img.youtube.com/vi/pyAu_vaIYZQ/maxresdefault.jpg)](https://www.youtube.com/watch?v=pyAu_vaIYZQ)
+[![Watch the full SuperLocalMemory demo: AI agent memory shared by Claude Code, Codex, ChatGPT and Grok Bot](https://img.youtube.com/vi/9Y0wAT07KTc/maxresdefault.jpg)](https://www.youtube.com/watch?v=9Y0wAT07KTc)
 
-*The four-minute product film: Claude Code and Codex sharing one memory.*
+*The full eight-minute demo: one memory shared by your coding agents and bots, every tab.*
 
 </div>
 
+## Add SuperLocalMemory to your app
+
+You do not need to wait for an app store. Every app below can take SuperLocalMemory straight from this GitHub repository, `qualixar/superlocalmemory`, in two steps.
+
+**Step 1, once per computer.** Install SuperLocalMemory: `npm install -g superlocalmemory` (or `pipx install superlocalmemory`). Your memory lives on this computer from then on.
+
+**Step 2, in each app you use:**
+
+| Your app | What to do |
+|---|---|
+| **Claude Code** (terminal and desktop app) | Type `/plugin marketplace add qualixar/superlocalmemory`, then `/plugin install superlocalmemory@qualixar` |
+| **Claude Desktop** | Run `slm connect claude-desktop`, then restart Claude Desktop |
+| **ChatGPT, Claude on the web, Muse, Composio** | Open the dashboard (`slm dashboard`) → **Web access** → turn it on, then add the link it shows as a connector in that app ([step-by-step](https://github.com/qualixar/superlocalmemory/blob/main/docs/remote-access/hosts.md)) |
+| **Codex** (terminal and desktop app) | Run `codex plugin marketplace add qualixar/superlocalmemory --ref main`, then `codex plugin add superlocalmemory-codex@qualixar` |
+| **Cursor** (editor and Agents window) | Run `slm connect cursor`, then restart Cursor |
+| **VS Code** (Copilot chat in the editor) | Install the SuperLocalMemory extension (`.vsix` from the [latest release](https://github.com/qualixar/superlocalmemory/releases/latest): Extensions → `...` → Install from VSIX) |
+| **VS Code Agents window, GitHub Copilot CLI** | Run `slm connect copilot-cli`; for the skills, add `qualixar/superlocalmemory` in the Agents window's Plugins screen |
+| **Antigravity** | Run `agy plugin install "$(npm root -g)/superlocalmemory/antigravity-plugin"` |
+| **Grok CLI** | Run `grok plugin install 'qualixar/superlocalmemory#plugin'` |
+| **Grok Bot** | In Grok's Plugins screen, add the marketplace `qualixar/superlocalmemory` and add `superlocalmemory` |
+| **Muse CLI** | Run `slm connect muse`, then `muse plugins install "$(npm root -g)/superlocalmemory/muse-plugin"` and `muse plugins approve superlocalmemory` |
+| **Hermes** (terminal and desktop app) | Run `hermes plugins install 'https://github.com/qualixar/superlocalmemory.git#hermes-plugin'`, then `slm upgrade-hosts --host hermes --apply` |
+| **OpenClaw** | Run `slm connect openclaw` |
+| **Windsurf, Zed, JetBrains, Gemini CLI, OpenCode, Continue** | Run `slm connect <app>`, for example `slm connect windsurf` |
+| **Anything else that speaks MCP** | Add a server with the command `slm` and the argument `mcp` |
+
+Restart the app after adding SuperLocalMemory. Then ask it something like *"What do you remember about this project?"* The [plugin guide](https://github.com/qualixar/superlocalmemory/blob/main/docs/plugins.md) has the details for each app, and `slm doctor` checks your setup.
+
 ## Works with your agents
 
 **Every harness, every bot, one memory.** Any harness that speaks MCP (`slm mcp`, or `http://127.0.0.1:8765/mcp/`) or can run a command (`slm remember`, `slm recall`) can use SuperLocalMemory. The packaged plugins are ready-made examples, not a limit.
@@ -147,7 +176,12 @@ SuperLocalMemory is part of Qualixar's AI Reliability Engineering work: agent me
 | **Codex** | `slm connect codex`, then `slm codex install` | MCP, 16 skills, 4 agents, `/slm-loop`, lifecycle hooks |
 | **GitHub Copilot** (VS Code, Visual Studio, JetBrains, CLI) | `slm connect vscode-copilot --here` | MCP, agent rules, 16 prompts, 4 agents, hooks |
 | **Antigravity** | `agy plugin install` the bundled plugin | MCP, 16 skills, 4 agents, `/slm-loop`, hooks |
-| **Hermes** | Native plugin | 16 skills, advisor agents, `/slm <command>` |
+| **Hermes** | Native plugin, then `slm upgrade-hosts --host hermes --apply` to grant it the memory server | 16 skills, advisor agents, `/slm <command>`, recall and session hooks |
+| **VS Code** | The SuperLocalMemory extension (`.vsix` on the [release page](https://github.com/qualixar/superlocalmemory/releases/latest)) | The memory server in the agent's tool list, nothing to edit |
+| **Grok CLI** | `slm connect grok-cli`, then `grok plugin install` the bundled plugin | MCP, 16 skills, 4 agents, session and recall hooks |
+| **VS Code Agents window, GitHub Copilot CLI** | `slm connect copilot-cli`; skills and agents from the Plugins screen | MCP, 16 skills, 4 agents |
+| **OpenClaw** | `slm connect openclaw` | MCP and 14 skills |
+| **Muse CLI** | `slm connect muse`, then `muse plugins install` the bundled plugin | MCP, 16 skills, 4 agents, `/slm-loop`, session and recall hooks |
 | **Grok Bot and Cursor-format hosts** | Add the `qualixar` plugin marketplace | A lean plugin built for shared bot computers ([guide](https://github.com/qualixar/superlocalmemory/blob/main/docs/plugins.md#grok-bot-and-other-cursor-format-hosts)) |
 | **Cursor, Windsurf, Zed, JetBrains, Gemini CLI, Claude Desktop, OpenCode, Continue** | `slm connect <name>` | MCP in that tool's own config |
 | **ChatGPT, Claude on the web, Muse, Composio** | Optional **Web access** from the dashboard: sign in per app, read-only or read-and-save, remove any app at once | Your local memory, from the web; the database never leaves your computer ([guide](https://github.com/qualixar/superlocalmemory/blob/main/docs/remote-access/README.md)) |
@@ -270,14 +304,21 @@ Engineering controls that support a compliance program, not a certification.
 
 ## MCP memory server: tool profiles
 
-Pick how many tools your agent sees with `SLM_MCP_PROFILE`.
+Every app, plugin and extension gets `full` (58 tools: all of memory, SLM-Mesh, optimize, loops) unless you choose otherwise. That fits hosts that cap the total number of MCP tools at about 100, with room for your other servers. To change it for every app at once, run one command and restart your apps:
+
+```bash
+slm tools                # current set, and how many tools each set has
+slm tools set power      # any name below; `slm tools set default` goes back to full
+```
+
+To change one app only, set `SLM_MCP_PROFILE` in that app's MCP config (`slm connect <app> --profile <name>` writes it); it wins over the saved choice for that app. [Per-app details](https://github.com/qualixar/superlocalmemory/blob/main/docs/plugins.md#choosing-how-many-tools)
 
 | Profile | Tools | For |
 |---|---:|---|
 | `core` | 19 | Remember, recall, expand, sessions, optimize, correction review |
 | `code` | 39 | Core plus code graph, memory kinds, Brain evidence, profile switching, bounded loops |
 | `mesh` | 9 | SLM-Mesh coordination only |
-| `full` (and unset) | 58 | Memory, kinds, Brain, optimize, skill evolution, mesh, loops, views and summaries (63 with pictures and documents on) |
+| `full` (default, when unset) | 58 | Memory, kinds, Brain, optimize, skill evolution, mesh, loops, views and summaries (63 with pictures and documents on) |
 | `power` | 70 | Full plus administration, lifecycle and diagnostics (75 with pictures and documents on) |
 | `whole` | 110 | Every registered tool |
 
@@ -291,7 +332,7 @@ stdio: `{"command": "slm", "args": ["mcp"]}` · HTTP: `http://127.0.0.1:8765/mcp
 | **Entity Explorer** | **The answer check at work** |
 | <img src="https://raw.githubusercontent.com/qualixar/superlocalmemory/main/docs/assets/gifs/entities.gif" alt="Entity Explorer: projects and tools with their facts and timeline"> | <img src="https://raw.githubusercontent.com/qualixar/superlocalmemory/main/docs/screenshots/dashboard/answer-check-answered.png" alt="Answer Check: the on-device model judges that the memories answer the question, with its confidence and timing"> |
 
-Recorded on a test install with fictional memories. [The four-minute product film](https://www.youtube.com/watch?v=pyAu_vaIYZQ) walks through every pane.
+Recorded on a test install with fictional memories. [The eight-minute full demo](https://www.youtube.com/watch?v=9Y0wAT07KTc) walks through every pane.
 
 ## Architecture
 

package/antigravity-plugin/_GENERATED.md

@@ -2,7 +2,7 @@
 
 Built by `scripts/build-antigravity-plugin.mjs` from `plugin-src/`. Version stamped from `plugin-src/manifest.json`.
 
-Version: **4.2.4**
+Version: **4.2.6**
 
 Do not edit by hand — regenerate instead.
 

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

@@ -81,4 +81,4 @@ slm-scope · slm-governance · slm-profile · slm-remember · slm-recall · slm-
 # What NOT to do
 Never session_init twice; never forget without dry-run preview; never run `slm gdpr erase --yes` unasked; never store secrets; never bypass role checks; never claim an erasure succeeded without verifying via recall.
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

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

@@ -79,4 +79,4 @@ assessment. The gate is the authority.
 
 ---
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

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

@@ -51,4 +51,4 @@ 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 switch the active profile unasked; 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 v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

package/antigravity-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 hand a ccr_id to another agent.
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

package/antigravity-plugin/mcp_config.json

@@ -6,8 +6,7 @@
         "mcp"
       ],
       "env": {
-        "SLM_AGENT_ID": "antigravity",
-        "SLM_MCP_PROFILE": "power"
+        "SLM_AGENT_ID": "antigravity"
       }
     }
   }

package/antigravity-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "superlocalmemory",
-  "version": "4.2.4",
+  "version": "4.2.6",
   "description": "Local-first agent memory with auditable hybrid retrieval. Remember decisions and recall them by asking, entirely on your machine.",
   "author": {
     "name": "Qualixar",

package/antigravity-plugin/skills/slm-bot-memory/SKILL.md

@@ -178,4 +178,4 @@ store every local bot on that computer can reach.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -144,4 +144,4 @@ These subcommands control daemon-level cache settings and bulk removal. They do
 
 ---
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

package/antigravity-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 v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

package/antigravity-plugin/skills/slm-getting-started-bot/SKILL.md

@@ -194,4 +194,4 @@ quality on a shared box.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -20,7 +20,9 @@ SuperLocalMemory can run with named users and roles per workspace (company
 mode), keeps a compliance audit trail, applies a retention lifecycle, and ships
 GDPR export and erasure commands. This skill documents how an agent must behave
 in a governed workspace and what the governance tools really do. The MCP audit
-and retention tools are in the `power` tool set (see `slm-profile`).
+and retention tools are in the `power` tool set, not the default: if they are
+missing, tell the user to run `slm tools set power` and restart this app (see
+`slm-profile`). The `slm gdpr` and `slm team` commands work either way.
 
 ---
 
@@ -274,4 +276,4 @@ Before running any destructive or state-changing operation (`forget`,
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -20,7 +20,7 @@ allowed-tools: build_code_graph, query_graph, get_blast_radius, semantic_search_
 
 # slm-graph — Code Intelligence Skill
 
-Index any repo as a code knowledge graph and answer structural questions about it: callers, callees, impact radius, semantic search, and review context. These tools are only registered in the `code` MCP tool set (or with `SLM_MCP_ALL_TOOLS=1`); see "Tool set requirement" below.
+Index any repo as a code knowledge graph and answer structural questions about it: callers, callees, impact radius, semantic search, and review context. These tools are not in the default tool set: they need `slm tools set code` (or `whole`); see "Tool set requirement" below.
 
 **Prerequisite rule:** every tool except `build_code_graph` self-guards — if the graph is not built it returns `{"success": false, "error": "Code graph not built. Run build_code_graph first."}`. Always index first.
 
@@ -289,17 +289,11 @@ If `build_code_graph` returns `files_parsed: 0`, no supported source files were
 The six tools above are registered only when the MCP server runs the `code` tool
 set, or with `SLM_MCP_ALL_TOOLS=1` / `SLM_MCP_PROFILE=whole`, or when
 `SLM_MCP_TOOLS` names them. They are **not** in `core`, `full`, `power` or the
-no-profile default that the Claude Code and Codex plugins use, so with those
-settings the tools simply do not exist and a call fails as an unknown tool. The
-tool set is read when the server starts: set it in the host's MCP config and
-restart the host.
-
-```json
-"env": {
-  "SLM_MCP_PROFILE": "code",
-  "SLM_AGENT_ID="antigravity"
-}
-```
+default every app gets, so there the tools do not exist and a call fails as an
+unknown tool. If the user asks for graph work and the tools are missing, tell
+them: "Code-graph tools are off by default. Run `slm tools set code` (graph
+without mesh) or `slm tools set whole` (every tool), then restart this app."
+Let them choose; it changes the tools every connected app sees.
 
 `switch_profile` does not change the tool set; it changes the active memory
 profile. See `slm-profile`. The `code` set leaves out the mesh and several
@@ -324,4 +318,4 @@ All are host-only: a remote caller cannot use them.
 
 ---
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

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

@@ -171,4 +171,4 @@ Guides: `docs/skill-evolution.md`, `docs/bounded-loops-bridge.md`.
 
 ---
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

package/antigravity-plugin/skills/slm-media/SKILL.md

@@ -337,4 +337,4 @@ minutes (previous vectors are kept) and let them decide.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -386,4 +386,4 @@ mesh availability.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -1,6 +1,6 @@
 ---
 name: slm-profile
-description: Memory profiles (separate namespaces inside one SuperLocalMemory store) and MCP tool sets (which tools your host sees). Use switch_profile or slm profile switch to change the active memory profile, the profile_id argument to read or write another profile for one call, and SLM_MCP_PROFILE to choose a tool set. Required when working across multiple projects, clients, or tenants.
+description: Memory profiles (separate namespaces inside one SuperLocalMemory store) and MCP tool sets (which tools your host sees). Use switch_profile or slm profile switch to change the active memory profile, the profile_id argument to read or write another profile for one call, and `slm tools set <name>` to choose the tool set every connected app sees. Required when working across multiple projects, clients, or tenants.
 when_to_use: |
   - "Switch to my work profile"
   - "What profile am I currently in?"
@@ -97,14 +97,15 @@ named differently (for example `superlocalmemory-personal` and
 
 | Profile | Tools | What it is |
 |---------|-------|------------|
-| `core` | 19 tools — remember, recall, search, fetch, expand, list_recent, update_memory, forget, session, optimize, corrections, summaries, switch_profile | Smallest set; the Grok Bot / Cursor plugin uses it |
+| `core` | 19 tools — remember, recall, search, fetch, expand, list_recent, update_memory, forget, session, optimize, corrections, summaries, switch_profile | Smallest set |
 | `code` | 39 tools — core + portable Brain evidence, report_outcome/report_feedback, 6 code-graph tools, memory kinds, bounded loops | For coding agents that need the graph; no mesh, no `get_status` |
-| `full` | 58 tools — everyday memory, delete_memory, get_status, observe, saved views, learning tools, skills, optimize, kinds, loops, mesh | Same set as the no-profile default |
+| `full` | 58 tools — everyday memory, delete_memory, get_status, observe, saved views, learning tools, skills, optimize, kinds, loops, mesh | The default for every app |
 | `power` | 70 tools — full + audit_trail, retention, compaction, consistency_check, behavioral and diagnostic tools | Governance and admin work |
 | `mesh` | 9 tools — mesh coordination only | Lightweight cross-session signalling |
 
-A host that sets no profile gets the 58-tool `full` set (this is what the Claude
-Code and Codex plugins do; the Antigravity plugin sets `power`). While pictures
+Every app gets the 58-tool `full` set unless the user chose another: no plugin
+or extension SuperLocalMemory ships sets one. It fits hosts that cap the total
+number of MCP tools at about 100. While pictures
 and documents are on (`slm media status`), `full` lists 62 tools and `power` 74:
 the five picture tools `remember_media`, `get_media`, `remember_document`,
 `media_status` and `media_upload_link` are added to both, and never to `core`,
@@ -117,19 +118,19 @@ such as `core_memory` and `settle_session_outcomes` exist only with
 `SLM_MCP_ALL_TOOLS=1`, `whole` or an explicit `SLM_MCP_TOOLS` list. Tools whose
 function is to manage the SLM computer are never available to a remote caller.
 
-The tool set is read when the MCP server starts. To change it, set it in the
-host's MCP config and restart the host; asking the server at runtime does not
-work:
+The tool set is read when the MCP server starts. To change it for every app at
+once, the user runs one command and then restarts their apps (or reconnects
+them in Dashboard → Connected apps); asking the server at runtime does not work:
 
-```json
-"env": {
-  "SLM_MCP_PROFILE": "code",
-  "SLM_AGENT_ID="antigravity"
-}
+```bash
+slm tools                 # the current set and how many tools each set has
+slm tools set power       # core, code, full, power, mesh or whole
+slm tools set default     # back to full
 ```
 
-`slm connect <ide> --profile <name>` writes that variable into a supported
-host's config. Setting `SLM_MCP_PROFILE` to a name that does not exist stops
+An app whose own MCP config sets `SLM_MCP_PROFILE` keeps that value: it wins
+over the saved choice, for that app only (`slm connect <ide> --profile <name>`
+writes it). Setting `SLM_MCP_PROFILE` to a name that does not exist stops
 `slm mcp` with a message listing the valid names. Older count-suffixed names
 such as `code20`, `full38` and `power50` still resolve, with a startup warning.
 On an install that already has memories, do not set `SLM_DATA_DIR` in the MCP
@@ -142,8 +143,10 @@ set. It is not in `mesh`.
 
 ## If a tool is missing
 
-Check the tool set first (`slm status` does not show it; look at the host's MCP
-config for `SLM_MCP_PROFILE`). Mesh tools need `full`, `power`, `mesh` or the
+Check the tool set first: `slm tools` shows it (an app's own `SLM_MCP_PROFILE`,
+if set, overrides it for that app). Tell the user which set has the tool and
+the command, for example "run `slm tools set power` and restart this app", and
+let them decide; widening the set adds tools to every connected app. Mesh tools need `full`, `power`, `mesh` or the
 default. Picture and PDF tools need `full`, `power` or `whole` and the feature
 turned on (`slm media enable`). Code-graph tools need `code` or `SLM_MCP_ALL_TOOLS=1`. Audit and
 retention tools need `power`. `report_outcome` and `report_feedback` need
@@ -161,4 +164,4 @@ anything but `core` or `mesh`.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -515,4 +515,4 @@ See `slm-profile`.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -372,4 +372,4 @@ sessions on this machine, so use it only when the user asks to move. See
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -180,4 +180,4 @@ to review the impact. See `slm-remember` for the full deletion discipline.
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -288,4 +288,4 @@ answered. To also surface shared or global facts, call `recall` with
 
 ---
 
-*SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later*
+*SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later*

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

@@ -228,4 +228,4 @@ the memory-engine upgrade is pending.
 
 ---
 
-SuperLocalMemory v4.2.4 · Qualixar · AGPL-3.0-or-later
+SuperLocalMemory v4.2.6 · Qualixar · AGPL-3.0-or-later

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

@@ -15,5 +15,5 @@
   "mcpServers": "./.mcp.json",
   "name": "superlocalmemory-codex",
   "repository": "https://github.com/qualixar/superlocalmemory",
-  "version": "4.2.4"
+  "version": "4.2.6"
 }