memstack-skill-loader 4.14.0__tar.gz → 4.15.0__tar.gz

memstack_skill_loader-4.15.0/PKG-INFO

@@ -0,0 +1,276 @@
+Metadata-Version: 2.4
+Name: memstack-skill-loader
+Version: 4.15.0
+Summary: MCP server that vector-indexes MemStack Pro skills for on-demand loading
+Project-URL: Homepage, https://memstack.pro
+Project-URL: Documentation, https://memstack.pro/docs
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+Requires-Dist: mcp<2.0.0,>=1.0.0
+Requires-Dist: lancedb>=0.6.0
+Requires-Dist: sentence-transformers>=2.2.0
+Requires-Dist: pyarrow>=14.0.0
+Requires-Dist: httpx>=0.24.0
+Requires-Dist: aiohttp>=3.9
+Provides-Extra: dev
+Requires-Dist: setuptools>=68; extra == "dev"
+
+# MemStack™ Skill Loader
+
+**130 skills for Claude Code:** 86 free + 44 Pro exclusive. Vector-indexed so CC loads only the skill it needs, saving your context window.
+
+## Quick Start (5 minutes)
+
+MemStack installs in two parts: the **skills** (via the Claude Code plugin marketplace) and the **engine** (this loader, via PyPI). You need both. Every command below is labeled **(in Claude Code)** or **(in terminal)** — running one in the wrong place is the most common setup mistake.
+
+**Step 1 — Install the free skills · (in Claude Code):**
+```
+/plugin marketplace add cwinvestments/memstack
+/plugin install memstack@cwinvestments-memstack
+```
+Run both commands. This unlocks the 86 free skills right away.
+
+> **SSH error?** ("Host key verification failed" on a fresh server that's never used GitHub over SSH.)
+> **Default fix · (in terminal)** — rewrite GitHub to HTTPS, then retry Step 1:
+> ```bash
+> git config --global url."https://github.com/".insteadOf "git@github.com:"
+> ```
+> **Backup fix · (in terminal)** — add GitHub's host key, then retry Step 1:
+> ```bash
+> mkdir -p ~/.ssh && ssh-keyscan github.com >> ~/.ssh/known_hosts
+> ```
+
+**Step 2 — Install the engine · (in terminal):**
+
+Windows:
+```bash
+pip install memstack-skill-loader
+```
+
+Linux / Mac — use the explicit interpreter (the same one you register in Step 3):
+```bash
+/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages
+```
+Find yours with `which python3` and substitute it if it isn't `/usr/bin/python3`. The `--break-system-packages` flag is required on newer externally-managed Python. No pip for that interpreter? Run `sudo apt install python3-pip` first (Debian/Ubuntu).
+
+**Step 3 — Register the MCP server · (in terminal):** register against the **same interpreter you installed onto in Step 2.**
+
+Windows:
+```bash
+claude mcp add --scope user memstack-skills -- python -m memstack_skill_loader
+```
+
+Linux / Mac:
+```bash
+claude mcp add --scope user memstack-skills -- /usr/bin/python3 -m memstack_skill_loader
+```
+If this doesn't match Step 2's interpreter, the server won't launch (you'll see a "failed to reconnect" error and `activate_license` will be missing).
+
+**Step 4 — Activate your license · (in Claude Code, after a full restart):** fully quit and reopen Claude Code first so it picks up the new MCP server, then run:
+```
+activate_license(key="your-key", email="you@example.com")
+```
+Use `key="free"` for the free tier, or your Pro key from [memstack.pro](https://memstack.pro) to unlock all 44 Pro skills (86 free + 44 Pro = 130 total). Then type `list skills` to verify.
+
+> **The first start is slow, and that is normal.** On the very first run after installing or upgrading, the server builds the skill index before it answers anything. Expect roughly a minute, nearly all of it loading the embedding model rather than reading your skills. **The server will look unresponsive for that time.** Let it finish. Every later start reuses the index and is not affected.
+
+> To override the skills path, set `MEMSTACK_SKILLS_DIR=/path/to/your/memstack/skills`.
+
+> See [Getting Started](https://memstack.pro/docs/getting-started) for detailed setup and [Skill Reference](https://memstack.pro/docs/skills) for the full skill catalog. If you hit issues, the [Troubleshooting](#troubleshooting) section below covers the common ones.
+
+## Troubleshooting
+
+**Nothing responds after a fresh install, and it does not seem to be failing either (during Step 4):** the first start builds the skill index before the server answers anything, which takes roughly a minute. Most of that is loading the embedding model, not reading your skills. There is no progress output, so a working first start and a hung one look identical. Wait it out before changing anything. If it is still silent well past a couple of minutes, then treat it as the interpreter mismatch below. Later starts reuse the index and respond immediately.
+
+**`error: externally-managed-environment` (during Step 2)** — pip refuses to install on a system-managed Python. Add `--break-system-packages` · (in terminal):
+```bash
+/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages
+```
+
+**Plugin clone fails: "Host key verification failed" (during Step 1)** — a fresh server that's never connected to GitHub over SSH. **Default fix · (in terminal):** `git config --global url."https://github.com/".insteadOf "git@github.com:"` then retry Step 1. **Backup fix · (in terminal):** `mkdir -p ~/.ssh && ssh-keyscan github.com >> ~/.ssh/known_hosts` then retry Step 1.
+
+**`activate_license` not found, or MCP "failed to reconnect (-32000)"** — the loader was installed onto a **different Python interpreter** than the one Claude Code launches (the classic bare-`python` mismatch on Linux/Mac). Check · (in terminal):
+```bash
+/usr/bin/python3 -c "import memstack_skill_loader; print('ok')"
+```
+If it prints `ok`, make sure Step 3 registered that exact path, then fully restart Claude Code. If it errors with `ModuleNotFoundError`, reinstall onto the explicit interpreter and re-register against that same path · (in terminal):
+```bash
+/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages
+claude mcp add --scope user memstack-skills -- /usr/bin/python3 -m memstack_skill_loader
+```
+Then fully restart Claude Code. Rule of thumb: whatever interpreter you `import`-check as `ok` is the path that must appear in your `claude mcp add` command.
+
+**`No module named pip` (during Step 2)** — that interpreter ships without pip (common on minimal Debian/Ubuntu server images). Install it, then retry the Step 2 install · (in terminal):
+```bash
+sudo apt install python3-pip
+```
+
+## How It Works
+
+MCP server that vector-indexes all 130 MemStack™ skills so Claude Code can call `find_skill("deploy to Railway")` and load **only** the relevant skill on demand, instead of all skills consuming context window.
+
+- **No Anthropic API key required:** skill search and indexing run locally and never leave your machine. Three small calls happen on their own at session start — license validation, a Pro-bundle version probe, and a PyPI update check — and every other outbound call is one you trigger ([Data and privacy](#data-and-privacy))
+- **Pro skills auto-detected:** once activated, the Pro skills directory is added to your sources automatically, with no `config.json` edit
+- **Auto-reindex on start:** skills stay current without manual rebuilds. Only startup ever rebuilds, never a search. The first start pays a one-time build of roughly a minute; after that a rebuild happens in the background while the existing index keeps serving, so later starts are not delayed
+- **Memory Engine:** an append-only, versioned fact store with query-time recall scoring — facts decay with age and decay more slowly the more they are recalled. Facts flow in from the diary's `## FACTS` block and are exposed to Claude through five `mem_*` tools (see [Available Tools](#available-tools))
+- **Living memory at session start:** the SessionStart hook injects a compact, token-budgeted digest of the project's top recall-scored facts into context. It fails open — a missing, empty, or unreachable store never blocks startup
+- **8-page localhost dashboard:** Overview, Skills Manager, Burn Report, Memory Browser (project-aware diary reader), Memory Inspector (read-only Memory Engine fact views), Agent Monitor, Referrals, and Settings
+
+### Data and privacy
+
+Outbound calls fall into two groups: the ones the loader makes on its own, and the ones you trigger. Both are listed in full.
+
+**Automatic — these happen at session start without you asking:**
+
+- **License validation** → POST to `admin.cwaffiliateinvestments.com/api/licenses/validate`, with:
+  - your **license key**
+  - a **machine ID** — a stable, hashed fingerprint that binds the license to your machine
+  - your **email**, if you provided one at activation
+  - your **environment** — OS name, Python version, and MemStack loader version
+
+  That is the complete payload.
+- **Pro bundle version probe** → GET to `admin.cwaffiliateinvestments.com/api/skills/pro-bundle/version`. Pro licenses only, at most once every 24 hours. Sends your license key as a bearer token and nothing else; it asks only whether a newer skill bundle exists.
+- **Update check** → GET to `pypi.org/pypi/memstack-skill-loader/json`. At most once every 24 hours. Sends no payload at all — it reads the latest published version number so the loader can tell you an upgrade is available.
+
+**Triggered by you — these never fire on their own:**
+
+- **Pro skills download** → GET to `admin.cwaffiliateinvestments.com/api/skills/pro-bundle`, when you run `activate_license` or `refresh_pro_skills`, and when the version probe above finds a newer bundle. Sends your license key.
+- **Agent Runner** → `api.anthropic.com`, only while you are running an agent. This is the one call that carries your actual work: the task you wrote and the contents of the files the agents read. It uses your own Anthropic API key, and nothing is sent unless you start an agent.
+- **Referrals page** → POST to `memstack.pro/api/referrals/me`, only when you open that dashboard page. Sends your license key.
+- **TokenStack proxy** → `api.anthropic.com`, only if you start the proxy yourself. It forwards your Claude Code traffic after compressing it, and is off unless you turn it on.
+
+Outside the Agent Runner and the TokenStack proxy — both of which you start deliberately, and both of which talk to Anthropic rather than to us — **no code, file contents, file paths, usernames, project names, or prompts are ever sent.** The environment fields in license validation exist only so platform-specific breakage (for example, a bug that affects only Linux) is visible to us in aggregate.
+
+### Environment Variable Override
+
+Set `MEMSTACK_SKILLS_DIR` to override the skills path in `config.json`:
+
+```bash
+export MEMSTACK_SKILLS_DIR=/path/to/your/memstack/skills
+```
+
+## Stack
+
+- Python 3.10+
+- [MCP SDK](https://pypi.org/project/mcp/) (stdio transport)
+- [LanceDB](https://lancedb.com/) (vector storage, zero external dependencies)
+- [sentence-transformers](https://www.sbert.net/) with `all-MiniLM-L6-v2` (384-dim local embeddings)
+
+## Available Tools
+
+### `find_skill`
+Search skills by describing what you need. Returns the most relevant skill(s) with full instructions.
+
+```
+find_skill(query="deploy to Railway", top_k=3)
+```
+
+### `list_skills`
+Browse the full skill catalog with names and descriptions.
+
+```
+list_skills()
+```
+
+### `get_skill`
+Load a specific skill by exact name.
+
+```
+get_skill(name="railway deploy")
+```
+
+### `reindex_skills`
+Rebuild the vector index after adding or modifying skills.
+
+```
+reindex_skills()
+```
+
+### Memory Engine
+
+The `mem_*` tools back a persistent, append-only fact store shared across sessions (see [How It Works](#how-it-works)).
+
+- **`mem_store`** — Store a new immutable fact (subject, claim, provenance). Returns the fact id and its provenance tag.
+- **`mem_correct`** — Correct a fact by superseding it with a new one. Corrections extend only from the live tip of a chain; correcting a superseded fact is refused and names the current tip.
+- **`mem_recall`** — Recall live facts for a project, each rendered with an inline provenance tag `[method · source · recalled Nx]`. Each returned fact is logged as a recall. Optional full-text query or exact-subject filter.
+- **`mem_inspect`** — Inspect a single fact and its full supersession chain (root → tip) with statuses, provenance, entities, and recall count.
+- **`mem_stats`** — Fact-store statistics: totals by status (live/superseded/retracted), distinct subjects, and recall count. Optionally scoped to a project.
+
+## Configuration
+
+The `config.json` file controls where skills are loaded from:
+
+```json
+{
+  "skill_sources": [
+    {
+      "type": "local",
+      "path": "/path/to/memstack/skills",
+      "pattern": "**/SKILL.md",
+      "label": "MemStack"
+    }
+  ],
+  "embedding_model": "all-MiniLM-L6-v2",
+  "default_top_k": 3,
+  "vector_db_path": "./vectors",
+  "auto_reindex_on_start": true
+}
+```
+
+`vector_db_path` sets where the skill index is stored. A relative value resolves against the canonical state root `~/.memstack`, not against the directory holding `config.json`, so the index lives in one place regardless of which directory a process is started from. An absolute or `~`-prefixed value is honored verbatim.
+
+You never add Pro skills to `config.json` by hand: the loader adds that directory itself once a key is present. Run `activate_license` at least once regardless, because that is the step that validates your key and downloads the Pro skills. `MEMSTACK_PRO_LICENSE_KEY` is an optional override that supplies the key without storing it, and it takes precedence over the stored one, but on its own it downloads nothing.
+
+Add entries to `skill_sources` to index skills from multiple directories:
+
+```json
+{
+  "skill_sources": [
+    {
+      "type": "local",
+      "path": "/path/to/memstack/skills",
+      "pattern": "**/SKILL.md",
+      "label": "MemStack"
+    },
+    {
+      "type": "local",
+      "path": "/home/user/custom-skills",
+      "pattern": "*.md",
+      "label": "My Custom Skills"
+    }
+  ]
+}
+```
+
+The `pattern` field controls how skills are discovered:
+- `**/SKILL.md`: Subdirectory structure (e.g., `category/skill-name/SKILL.md`)
+- `*.md`: Flat directory (each `.md` file is a skill)
+
+## Release Notes
+
+### 4.15.0 (August 2026)
+
+- The skill index moved to the canonical state root, so a relative `vector_db_path` no longer depends on which directory the server happened to start in. **The first start after installing or upgrading rebuilds the index once, which takes roughly a minute**, nearly all of it loading the embedding model. Later starts are unaffected.
+- New index guard classifies the index as current, drift, stale, absent, or rebuilding, and reports that verdict at startup and in the dashboard. Only startup may rebuild on its own, so no search request ever blocks on one.
+- Agent Runner git policy rewritten as an argv-level classifier, replacing a substring match that could not tell `git status` from `git push`.
+
+### 4.10.0 to 4.14.1 (July 2026)
+
+- Memory Engine: diary `## FACTS` ingestion, living-memory injection at session start, a read-only Memory Inspector dashboard page, and recall that ranks the whole live fact set rather than a window of it.
+- Added the goal skill, bringing the public catalog to 130 skills (86 free + 44 Pro).
+- Added the video-review Pro skill.
+- Referrals page added to the dashboard.
+- License validation now discloses its full payload in-product (see [Data and privacy](#data-and-privacy)).
+- Packaging fix: `skill_descriptions.json` now ships in the wheel, so a pip install no longer falls back to empty dashboard descriptions.
+
+### 4.0.0 to 4.9.0 (May to July 2026)
+
+- Dashboard and Agent Runner: 3-agent orchestration (Manager, Builder, Reviewer) with per-agent model selection, context window monitoring, and token cost tracking.
+- Session diary with markdown narratives, and safe git staging that keeps secrets and runtime data out of commits.
+- TokenStack proxy compression, including AST compression of line-numbered Python.
+- Memory Browser with project filtering and secret masking.
+- Licensing hardened against transient network failures and machine mismatches.
+
+## License
+
+Proprietary. Part of MemStack™ Pro by CW Affiliate Investments LLC.

{memstack_skill_loader-4.14.0 → memstack_skill_loader-4.15.0}/README.md

@@ -55,12 +55,16 @@ activate_license(key="your-key", email="you@example.com")
 ```
 Use `key="free"` for the free tier, or your Pro key from [memstack.pro](https://memstack.pro) to unlock all 44 Pro skills (86 free + 44 Pro = 130 total). Then type `list skills` to verify.
 
+> **The first start is slow, and that is normal.** On the very first run after installing or upgrading, the server builds the skill index before it answers anything. Expect roughly a minute, nearly all of it loading the embedding model rather than reading your skills. **The server will look unresponsive for that time.** Let it finish. Every later start reuses the index and is not affected.
+
 > To override the skills path, set `MEMSTACK_SKILLS_DIR=/path/to/your/memstack/skills`.
 
-> See [QUICKSTART.md](QUICKSTART.md) for detailed setup, [QUICK-REFERENCE.md](QUICK-REFERENCE.md) for the full skill catalog, and [TROUBLESHOOTING.md](TROUBLESHOOTING.md) if you hit issues.
+> See [Getting Started](https://memstack.pro/docs/getting-started) for detailed setup and [Skill Reference](https://memstack.pro/docs/skills) for the full skill catalog. If you hit issues, the [Troubleshooting](#troubleshooting) section below covers the common ones.
 
 ## Troubleshooting
 
+**Nothing responds after a fresh install, and it does not seem to be failing either (during Step 4):** the first start builds the skill index before the server answers anything, which takes roughly a minute. Most of that is loading the embedding model, not reading your skills. There is no progress output, so a working first start and a hung one look identical. Wait it out before changing anything. If it is still silent well past a couple of minutes, then treat it as the interpreter mismatch below. Later starts reuse the index and respond immediately.
+
 **`error: externally-managed-environment` (during Step 2)** — pip refuses to install on a system-managed Python. Add `--break-system-packages` · (in terminal):
 ```bash
 /usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages
@@ -88,23 +92,37 @@ sudo apt install python3-pip
 
 MCP server that vector-indexes all 130 MemStack™ skills so Claude Code can call `find_skill("deploy to Railway")` and load **only** the relevant skill on demand, instead of all skills consuming context window.
 
-- **No Anthropic API key required:** skill search and indexing run locally — the only outbound network call is license validation, which contacts the MemStack license server on session start ([Data and privacy](#data-and-privacy))
-- **Pro skills auto-detected:** set your license key and they appear automatically
-- **Auto-reindex on start:** skills stay current without manual rebuilds
+- **No Anthropic API key required:** skill search and indexing run locally and never leave your machine. Three small calls happen on their own at session start — license validation, a Pro-bundle version probe, and a PyPI update check — and every other outbound call is one you trigger ([Data and privacy](#data-and-privacy))
+- **Pro skills auto-detected:** once activated, the Pro skills directory is added to your sources automatically, with no `config.json` edit
+- **Auto-reindex on start:** skills stay current without manual rebuilds. Only startup ever rebuilds, never a search. The first start pays a one-time build of roughly a minute; after that a rebuild happens in the background while the existing index keeps serving, so later starts are not delayed
 - **Memory Engine:** an append-only, versioned fact store with query-time recall scoring — facts decay with age and decay more slowly the more they are recalled. Facts flow in from the diary's `## FACTS` block and are exposed to Claude through five `mem_*` tools (see [Available Tools](#available-tools))
 - **Living memory at session start:** the SessionStart hook injects a compact, token-budgeted digest of the project's top recall-scored facts into context. It fails open — a missing, empty, or unreachable store never blocks startup
 - **8-page localhost dashboard:** Overview, Skills Manager, Burn Report, Memory Browser (project-aware diary reader), Memory Inspector (read-only Memory Engine fact views), Agent Monitor, Referrals, and Settings
 
 ### Data and privacy
 
-License validation is the loader's only outbound network call. On session start it POSTs to the MemStack license server (`admin.cwaffiliateinvestments.com/api/licenses/validate`) with:
+Outbound calls fall into two groups: the ones the loader makes on its own, and the ones you trigger. Both are listed in full.
+
+**Automatic — these happen at session start without you asking:**
+
+- **License validation** → POST to `admin.cwaffiliateinvestments.com/api/licenses/validate`, with:
+  - your **license key**
+  - a **machine ID** — a stable, hashed fingerprint that binds the license to your machine
+  - your **email**, if you provided one at activation
+  - your **environment** — OS name, Python version, and MemStack loader version
+
+  That is the complete payload.
+- **Pro bundle version probe** → GET to `admin.cwaffiliateinvestments.com/api/skills/pro-bundle/version`. Pro licenses only, at most once every 24 hours. Sends your license key as a bearer token and nothing else; it asks only whether a newer skill bundle exists.
+- **Update check** → GET to `pypi.org/pypi/memstack-skill-loader/json`. At most once every 24 hours. Sends no payload at all — it reads the latest published version number so the loader can tell you an upgrade is available.
 
-- your **license key**
-- a **machine ID** — a stable, hashed fingerprint that binds the license to your machine
-- your **email**, if you provided one at activation
-- your **environment** — OS name, Python version, and MemStack loader version
+**Triggered by you — these never fire on their own:**
 
-That is the complete payload. **No code, file contents, file paths, usernames, project names, or prompts are ever sent.** The environment fields exist only so platform-specific breakage (for example, a bug that affects only Linux) is visible to us in aggregate.
+- **Pro skills download** → GET to `admin.cwaffiliateinvestments.com/api/skills/pro-bundle`, when you run `activate_license` or `refresh_pro_skills`, and when the version probe above finds a newer bundle. Sends your license key.
+- **Agent Runner** → `api.anthropic.com`, only while you are running an agent. This is the one call that carries your actual work: the task you wrote and the contents of the files the agents read. It uses your own Anthropic API key, and nothing is sent unless you start an agent.
+- **Referrals page** → POST to `memstack.pro/api/referrals/me`, only when you open that dashboard page. Sends your license key.
+- **TokenStack proxy** → `api.anthropic.com`, only if you start the proxy yourself. It forwards your Claude Code traffic after compressing it, and is off unless you turn it on.
+
+Outside the Agent Runner and the TokenStack proxy — both of which you start deliberately, and both of which talk to Anthropic rather than to us — **no code, file contents, file paths, usernames, project names, or prompts are ever sent.** The environment fields in license validation exist only so platform-specific breakage (for example, a bug that affects only Linux) is visible to us in aggregate.
 
 ### Environment Variable Override
 
@@ -116,7 +134,7 @@ export MEMSTACK_SKILLS_DIR=/path/to/your/memstack/skills
 
 ## Stack
 
-- Python 3.12+
+- Python 3.10+
 - [MCP SDK](https://pypi.org/project/mcp/) (stdio transport)
 - [LanceDB](https://lancedb.com/) (vector storage, zero external dependencies)
 - [sentence-transformers](https://www.sbert.net/) with `all-MiniLM-L6-v2` (384-dim local embeddings)
@@ -170,7 +188,7 @@ The `config.json` file controls where skills are loaded from:
   "skill_sources": [
     {
       "type": "local",
-      "path": "C:\\Projects\\memstack\\skills",
+      "path": "/path/to/memstack/skills",
       "pattern": "**/SKILL.md",
       "label": "MemStack"
     }
@@ -182,7 +200,9 @@ The `config.json` file controls where skills are loaded from:
 }
 ```
 
-Pro skills are **auto-detected** when `MEMSTACK_PRO_LICENSE_KEY` is set, no need to add them to `config.json`.
+`vector_db_path` sets where the skill index is stored. A relative value resolves against the canonical state root `~/.memstack`, not against the directory holding `config.json`, so the index lives in one place regardless of which directory a process is started from. An absolute or `~`-prefixed value is honored verbatim.
+
+You never add Pro skills to `config.json` by hand: the loader adds that directory itself once a key is present. Run `activate_license` at least once regardless, because that is the step that validates your key and downloads the Pro skills. `MEMSTACK_PRO_LICENSE_KEY` is an optional override that supplies the key without storing it, and it takes precedence over the stored one, but on its own it downloads nothing.
 
 Add entries to `skill_sources` to index skills from multiple directories:
 
@@ -191,7 +211,7 @@ Add entries to `skill_sources` to index skills from multiple directories:
   "skill_sources": [
     {
       "type": "local",
-      "path": "C:\\Projects\\memstack\\skills",
+      "path": "/path/to/memstack/skills",
       "pattern": "**/SKILL.md",
       "label": "MemStack"
     },
@@ -211,20 +231,28 @@ The `pattern` field controls how skills are discovered:
 
 ## Release Notes
 
-### v4.0.0 (May 2026)
+### 4.15.0 (August 2026)
+
+- The skill index moved to the canonical state root, so a relative `vector_db_path` no longer depends on which directory the server happened to start in. **The first start after installing or upgrading rebuilds the index once, which takes roughly a minute**, nearly all of it loading the embedding model. Later starts are unaffected.
+- New index guard classifies the index as current, drift, stale, absent, or rebuilding, and reports that verdict at startup and in the dashboard. Only startup may rebuild on its own, so no search request ever blocks on one.
+- Agent Runner git policy rewritten as an argv-level classifier, replacing a substring match that could not tell `git status` from `git push`.
+
+### 4.10.0 to 4.14.1 (July 2026)
+
+- Memory Engine: diary `## FACTS` ingestion, living-memory injection at session start, a read-only Memory Inspector dashboard page, and recall that ranks the whole live fact set rather than a window of it.
+- Added the goal skill, bringing the public catalog to 130 skills (86 free + 44 Pro).
+- Added the video-review Pro skill.
+- Referrals page added to the dashboard.
+- License validation now discloses its full payload in-product (see [Data and privacy](#data-and-privacy)).
+- Packaging fix: `skill_descriptions.json` now ships in the wheel, so a pip install no longer falls back to empty dashboard descriptions.
 
-- Dashboard: 6-page localhost dashboard (Overview, Skills Manager, Burn Report, Memory Browser, Agent Monitor, Settings)
-- Agent Runner: 3-agent orchestration (Manager/Builder/Reviewer) with per-agent model selection
-- 17 MCP tools
-- 127 skills (84 free + 43 Pro)
-- Real-time context window monitoring per agent
-- Session diary with AI-authored markdown narratives
-- Safe git staging (prevents accidental commits of secrets/runtime data)
-- Task completion notifications (browser, tab flash, audio)
-- Token usage tracking with estimated costs
-- TokenStack™ proxy integration (~35-40% token savings)
+### 4.0.0 to 4.9.0 (May to July 2026)
 
-**[v3.4.0](https://github.com/cwinvestments/memstack-skill-loader/releases/tag/v3.4.0)**: 100 Skills Milestone (18 new Pro skills, auto-detection, display name fixes)
+- Dashboard and Agent Runner: 3-agent orchestration (Manager, Builder, Reviewer) with per-agent model selection, context window monitoring, and token cost tracking.
+- Session diary with markdown narratives, and safe git staging that keeps secrets and runtime data out of commits.
+- TokenStack proxy compression, including AST compression of line-numbered Python.
+- Memory Browser with project filtering and secret masking.
+- Licensing hardened against transient network failures and machine mismatches.
 
 ## License
 

memstack_skill_loader-4.15.0/pyproject.toml

@@ -0,0 +1,51 @@
+[build-system]
+requires = ["setuptools>=68.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "memstack-skill-loader"
+version = "4.15.0"
+description = "MCP server that vector-indexes MemStack Pro skills for on-demand loading"
+readme = "README.md"
+requires-python = ">=3.10"
+dependencies = [
+    # Upper bound is load-bearing: mcp 2.0.0 (2026-07-28) removed the
+    # Server.list_tools()/call_tool() decorator API that server.py is built on.
+    # Do not relax without porting server.py to the 2.x API first.
+    "mcp>=1.0.0,<2.0.0",
+    "lancedb>=0.6.0",
+    "sentence-transformers>=2.2.0",
+    "pyarrow>=14.0.0",
+    "httpx>=0.24.0",
+    "aiohttp>=3.9",
+]
+
+[project.urls]
+Homepage = "https://memstack.pro"
+Documentation = "https://memstack.pro/docs"
+
+[project.optional-dependencies]
+# setuptools is a TEST dependency here, not only a build dependency. The
+# packaging guard in tests/test_wheel_contents.py builds a wheel and inspects
+# it, and it fails rather than skips when the backend is absent, because a skip
+# is a silent pass. Python 3.12+ venvs no longer ship setuptools, so a bare venv
+# needs:  pip install -e ".[dev]"
+dev = ["setuptools>=68"]
+
+[project.scripts]
+memstack-index = "memstack_skill_loader.indexer:main"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.setuptools.package-data]
+memstack_skill_loader = ["dashboard.html", "skill_descriptions.json"]
+
+[tool.pytest.ini_options]
+# Test the WORKING TREE, not whatever is installed. This package is installed
+# non-editable, so without this the suite imported memstack_skill_loader from
+# site-packages and silently graded a stale copy: a source change could be
+# green here and still be broken, and a new test could fail against code that
+# was already fixed. pythonpath puts src/ ahead of site-packages for the whole
+# suite, so `python -m pytest` needs no PYTHONPATH prefix to be correct.
+pythonpath = ["src"]

{memstack_skill_loader-4.14.0 → memstack_skill_loader-4.15.0}/src/memstack_skill_loader/__init__.py

@@ -1,3 +1,3 @@
 """MemStack Skill Loader — MCP server for semantic skill search."""
 
-__version__ = "4.14.0"
+__version__ = "4.15.0"

{memstack_skill_loader-4.14.0 → memstack_skill_loader-4.15.0}/src/memstack_skill_loader/agent_runner.py

@@ -24,6 +24,9 @@ from .stats import log_agent_invocation, log_api_cost
 # ---------------------------------------------------------------------------
 
 MODEL_PRICING = {
+    "claude-fable-5": {"input": 10.0, "output": 50.0},
+    "claude-opus-5": {"input": 5.0, "output": 25.0},
+    "claude-sonnet-5": {"input": 3.0, "output": 15.0},
     "claude-sonnet-4-6": {"input": 3.0, "output": 15.0},
     "claude-opus-4-6": {"input": 5.0, "output": 25.0},
     "claude-opus-4-7": {"input": 5.0, "output": 25.0},
@@ -177,7 +180,15 @@ SYSTEM_PROMPTS = {
         "- Never hardcode secrets, API keys, tokens, passwords, or credentials in source code. "
         "Always use environment variables loaded from .env files. If a secret is needed, create "
         "or update a .env.example file with the variable name and a placeholder value.\n"
-        "- Do NOT run git add, git commit, or any git commands. Only make file changes."
+        "- Read-only git IS available to you: status, log, diff, show, ls-files, "
+        "check-ignore, branch/remote listing, ls-remote, rev-parse. Use it to "
+        "understand the repository before you change it.\n"
+        "- Do NOT run git add, git commit, git push, git merge, git checkout, or "
+        "git reset. Committing is the orchestrator's job, not yours. These are "
+        "denied in code, not on the honour system, so do not try to work around "
+        "the denial.\n"
+        "- Issue ONE git command per Bash call. Chaining with && or ; is denied, "
+        "as are output redirection and `git -C <other-repo>`."
     ),
     "advisor": (
         "You are an Advisor agent. You are called in ONLY when a Builder agent has failed "
@@ -214,6 +225,11 @@ SYSTEM_PROMPTS = {
         "verdict that the orchestrator parses as a hard contract.\n\n"
         "You MUST read the actual changed files to verify correctness. Do not trust the Builder's "
         "summary alone.\n\n"
+        "Read-only git is available to you for verification: status, log, diff, show, ls-files, "
+        "branch/remote listing, ls-remote, rev-parse, symbolic-ref. Git writes (commit, push, "
+        "merge, checkout, reset, fetch) are denied in code. Issue ONE git command per Bash call: "
+        "chaining with && or ;, output redirection, and `git -C <other-repo>` are all denied, so "
+        "run them as separate calls rather than trying to combine them.\n\n"
         "## Review Process\n\n"
         "Evaluate the submission across these 4 perspectives:\n"
         "1. **SECURITY** — secret/credential exposure, injection risks, auth/permission changes, "
@@ -1062,6 +1078,19 @@ def _build_env(strip_api_key: bool = False, inject_api_key: str | None = None) -
         env["ANTHROPIC_API_KEY"] = inject_api_key
     if ANTHROPIC_BASE_URL:
         env["ANTHROPIC_BASE_URL"] = ANTHROPIC_BASE_URL
+    # Git hardening for agent subprocesses. A PreToolUse hook can only allow or
+    # deny, never rewrite, so the vectors that live in repo-local config are
+    # closed here instead (AGENT_GIT_CAPABILITY_DESIGN.md §1.2):
+    #   GIT_PAGER          - .git/config core.pager is arbitrary code execution
+    #   GIT_EXTERNAL_DIFF  - .gitattributes can point `git diff` at a program
+    #   GIT_CONFIG_NOSYSTEM- ignore system-wide config the agent could reach
+    #   GIT_TERMINAL_PROMPT- a network read against an unauthenticated remote
+    #                        must error, not hang until the inactivity watchdog
+    env["GIT_PAGER"] = "cat"
+    env["GIT_EXTERNAL_DIFF"] = ""
+    env["GIT_CONFIG_NOSYSTEM"] = "1"
+    env["GIT_TERMINAL_PROMPT"] = "0"
+    env["GIT_ASKPASS"] = ""
     return env
 
 
@@ -1764,13 +1793,33 @@ def _parse_subtasks(text: str) -> list[tuple[str, str, list[str]]]:
 
 
 def _build_builder_hook_settings() -> str:
-    """Inline --settings JSON that registers the Builder git-deny PreToolUse hook.
+    """Inline --settings JSON that registers the Builder git PreToolUse hook.
 
     The hook script ships inside this package (hooks/deny_git.py). Its absolute
     path MUST use forward slashes: backslashes get stripped when the settings
     command string is executed on Windows, which mangles the path and breaks
-    the hook (see smoke test)."""
+    the hook (see smoke test).
+
+    The matcher covers the file-writing tools as well as Bash because the hook
+    now also protects .gitattributes / .gitconfig / .git/** from being authored
+    (AGENT_GIT_CAPABILITY_DESIGN.md §1.3); a Bash-only matcher would leave the
+    Write/Edit route wide open."""
     hook_script = os.path.join(os.path.dirname(__file__), "hooks", "deny_git.py").replace("\\", "/")
+    return json.dumps({
+        "hooks": {"PreToolUse": [{"matcher": "Write|Edit|MultiEdit|NotebookEdit|Bash",
+            "hooks": [{"type": "command", "command": f"python {hook_script}"}]}]}
+    })
+
+
+def _build_reviewer_hook_settings() -> str:
+    """Inline --settings JSON that registers the Reviewer read-only-git hook.
+
+    Mirrors the other two builders, including the forward-slash path (Windows
+    strips backslashes out of the settings command string). The Reviewer had no
+    PreToolUse hook at all before this; Bash is the only matcher because the
+    Reviewer must stay free to read files, build, and test in order to verify
+    a deploy independently."""
+    hook_script = os.path.join(os.path.dirname(__file__), "hooks", "deny_reviewer_git.py").replace("\\", "/")
     return json.dumps({
         "hooks": {"PreToolUse": [{"matcher": "Bash",
             "hooks": [{"type": "command", "command": f"python {hook_script}"}]}]}
@@ -2034,6 +2083,9 @@ def _orchestrate(session: Session) -> None:
     # Manager-only read-only hook (Layer 1). Passed to the Manager
     # _invoke_agent call via extra_settings; never to Builder/Reviewer.
     manager_hook_settings = _build_manager_hook_settings()
+    # Reviewer-only read-only-git hook (Layer 1). Passed to both Reviewer
+    # _invoke_agent call sites (subtask and single-task paths).
+    reviewer_hook_settings = _build_reviewer_hook_settings()
 
     try:
         # Load execution modes and custom display names for this run
@@ -2356,6 +2408,7 @@ def _orchestrate(session: Session) -> None:
                             model=session.models.get("reviewer", ""),
                             display_name=display_names["reviewer"],
                             execution_mode=agent_modes["reviewer"],
+                            extra_settings=reviewer_hook_settings,
                         )
                     try:
                         # Gap 2: strict verdict contract, fail-closed, re-prompt at most once.
@@ -2475,6 +2528,12 @@ def _orchestrate(session: Session) -> None:
                         _re.compile(r'__pycache__[\\/]'),
                         _re.compile(r'node_modules[\\/]'),
                         _re.compile(r'^\.env'),
+                        # Repository config: never auto-staged. A .gitattributes
+                        # or .git/config change can turn read-only git into code
+                        # execution (AGENT_GIT_CAPABILITY_DESIGN.md §1.3).
+                        _re.compile(r'(^|[\\/])\.gitattributes$'),
+                        _re.compile(r'(^|[\\/])\.gitconfig$'),
+                        _re.compile(r'(^|[\\/])\.git[\\/]'),
                         _re.compile(r'secret', _re.IGNORECASE),
                         _re.compile(r'credential', _re.IGNORECASE),
                     ]
@@ -2730,6 +2789,7 @@ def _orchestrate(session: Session) -> None:
                     model=session.models.get("reviewer", ""),
                     display_name=display_names["reviewer"],
                     execution_mode=agent_modes["reviewer"],
+                    extra_settings=reviewer_hook_settings,
                 )
             try:
                 # Gap 2: strict verdict contract, fail-closed, re-prompt at most once.
@@ -2793,6 +2853,12 @@ def _orchestrate(session: Session) -> None:
                             _re.compile(r'__pycache__[\\/]'),
                             _re.compile(r'node_modules[\\/]'),
                             _re.compile(r'^\.env'),
+                            # Repository config: never auto-staged. A .gitattributes
+                            # or .git/config change can turn read-only git into
+                            # code execution (AGENT_GIT_CAPABILITY_DESIGN.md §1.3).
+                            _re.compile(r'(^|[\\/])\.gitattributes$'),
+                            _re.compile(r'(^|[\\/])\.gitconfig$'),
+                            _re.compile(r'(^|[\\/])\.git[\\/]'),
                             _re.compile(r'secret', _re.IGNORECASE),
                             _re.compile(r'credential', _re.IGNORECASE),
                         ]

{memstack_skill_loader-4.14.0 → memstack_skill_loader-4.15.0}/src/memstack_skill_loader/categories.py

@@ -18,7 +18,6 @@ CATEGORY_MAP = {
     "humanize": "Content", "blog-post": "Content", "email-sequence": "Content",
     "landing-page-copy": "Content", "newsletter": "Content", "product-description": "Content",
     "tiktok-script": "Content", "twitter-thread": "Content", "youtube-script": "Content",
-    "kdp-format": "Content",
     # Deployment
     "railway-deploy": "Deployment", "docker-setup": "Deployment", "netlify-deploy": "Deployment",
     "domain-ssl": "Deployment", "hetzner-setup": "Deployment", "ci-cd-pipeline": "Deployment",