Rebrand to memxt (Yupcha/memxt)

.claude-plugin/marketplace.json

@@ -1,15 +1,15 @@
 {
-  "name": "mempalace",
-  "owner": { "name": "MemPalace", "url": "https://github.com/debpalash" },
+  "name": "memxt",
+  "owner": { "name": "Yupcha", "url": "https://github.com/Yupcha" },
   "metadata": {
     "description": "Local-first memory for AI coding agents.",
     "version": "0.2.0"
   },
   "plugins": [
     {
-      "name": "fast-mempalace",
+      "name": "memxt",
       "source": "./claude-plugin",
-      "description": "Persistent, private, local memory for Claude Code — your agent remembers your code and decisions across sessions. One static binary, no cloud. Requires the fast-mempalace binary (curl install).",
+      "description": "Persistent, private, local memory for Claude Code — your agent remembers your code and decisions across sessions. One static binary, no cloud. Requires the memxt binary (curl install).",
       "version": "0.2.0",
       "license": "MIT",
       "keywords": ["memory", "local-first", "mcp", "privacy"]

.github/workflows/build.yml

@@ -119,11 +119,11 @@ jobs:
       - name: Execute CLI Validation (incl. real embeddings)
         run: |
           set -eux
-          ./zig-out/bin/fast-mempalace init
-          ./zig-out/bin/fast-mempalace stats
+          ./zig-out/bin/memxt init
+          ./zig-out/bin/memxt stats
           printf 'The deploy key id is QUASAR-77 and must never be logged.\n' > smoke.txt
-          ./zig-out/bin/fast-mempalace mine smoke.txt smoke
-          ./zig-out/bin/fast-mempalace search "what is the deploy key" 2>&1 | grep -q QUASAR-77
+          ./zig-out/bin/memxt mine smoke.txt smoke
+          ./zig-out/bin/memxt search "what is the deploy key" 2>&1 | grep -q QUASAR-77
 
       - name: Integration tests (mine / ranking / MCP / hooks / schema)
-        run: bash scripts/integration-test.sh ./zig-out/bin/fast-mempalace
+        run: bash scripts/integration-test.sh ./zig-out/bin/memxt

.github/workflows/release.yml

@@ -115,29 +115,29 @@ jobs:
       - name: Smoke test (incl. real embeddings)
         run: |
           set -eux
-          ./zig-out/bin/fast-mempalace init
-          ./zig-out/bin/fast-mempalace stats
+          ./zig-out/bin/memxt init
+          ./zig-out/bin/memxt stats
           # Validate the full semantic path: mine a file, then recall it.
           printf 'The deploy key id is QUASAR-77 and must never be logged.\n' > smoke.txt
-          ./zig-out/bin/fast-mempalace mine smoke.txt smoke
-          ./zig-out/bin/fast-mempalace search "what is the deploy key" 2>&1 | grep -q QUASAR-77
+          ./zig-out/bin/memxt mine smoke.txt smoke
+          ./zig-out/bin/memxt search "what is the deploy key" 2>&1 | grep -q QUASAR-77
 
       - name: Package tarball
         env:
           TARGET: ${{ matrix.target }}
         run: |
-          TARBALL="fast-mempalace-${TARGET}.tar.gz"
-          tar -czf "$TARBALL" -C zig-out/bin fast-mempalace
+          TARBALL="memxt-${TARGET}.tar.gz"
+          tar -czf "$TARBALL" -C zig-out/bin memxt
           shasum -a 256 "$TARBALL" > "$TARBALL.sha256"
           ls -lh "$TARBALL" "$TARBALL.sha256"
 
       - name: Upload artifact
         uses: actions/upload-artifact@v4
         with:
-          name: fast-mempalace-${{ matrix.target }}
+          name: memxt-${{ matrix.target }}
           path: |
-            fast-mempalace-${{ matrix.target }}.tar.gz
-            fast-mempalace-${{ matrix.target }}.tar.gz.sha256
+            memxt-${{ matrix.target }}.tar.gz
+            memxt-${{ matrix.target }}.tar.gz.sha256
 
   release:
     name: Publish GitHub Release
@@ -164,7 +164,7 @@ jobs:
             Install with one line:
 
             ```bash
-            curl -fsSL https://raw.githubusercontent.com/MemPalace/fast-mempalace/main/install.sh | bash
+            curl -fsSL https://raw.githubusercontent.com/Yupcha/memxt/main/install.sh | bash
             ```
 
             Prebuilt binaries attached for darwin-aarch64, darwin-x86_64, linux-x86_64, linux-aarch64.

.gitignore

@@ -4,7 +4,7 @@ zig-out/
 .zig-cache/
 
 # Executables
-fast-mempalace
+memxt
 mempalace
 
 # IDE / OS
@@ -27,7 +27,7 @@ lib/*.gguf
 lib/llama.cpp/build/
 
 # External Configs mapping over default parameters
-fast-mempalace.yaml
+memxt.yaml
 mempalace.yaml
 lib/llama.cpp
 .claude/

BENCHMARK.md

@@ -1,4 +1,4 @@
-# ⏱️ fast-mempalace Benchmarks
+# ⏱️ memxt Benchmarks
 
 Absolute performance of the real semantic engine (`llama.cpp` MiniLM-L6-v2 embeddings +
 `sqlite-vec`). Hardware: Apple Silicon, macOS, Metal backend. Each command run cold
@@ -47,14 +47,14 @@ unrelated topics ~0.03.
 zig build --release=fast
 
 # point at a 384-dim model
-export FAST_MEMPALACE_MODEL=$PWD/lib/minilm.gguf
-export FAST_MEMPALACE_DB=/tmp/bench.db
+export MEMXT_MODEL=$PWD/lib/minilm.gguf
+export MEMXT_DB=/tmp/bench.db
 
-./zig-out/bin/fast-mempalace init
-/usr/bin/time -p ./zig-out/bin/fast-mempalace mine ./src bench     # mine a directory
-/usr/bin/time -p ./zig-out/bin/fast-mempalace search "how does search ranking work"
-/usr/bin/time -p ./zig-out/bin/fast-mempalace wake-up
+./zig-out/bin/memxt init
+/usr/bin/time -p ./zig-out/bin/memxt mine ./src bench     # mine a directory
+/usr/bin/time -p ./zig-out/bin/memxt search "how does search ranking work"
+/usr/bin/time -p ./zig-out/bin/memxt wake-up
 ```
 
-For warm MCP-server latency, start `fast-mempalace mcp` and issue repeated
+For warm MCP-server latency, start `memxt mcp` and issue repeated
 `tools/call → memory_search` requests over stdio; only the first pays the model-load cost.

Dockerfile.ci

@@ -1,8 +1,8 @@
 # Dockerfile.ci — local mirror of the Ubuntu GitHub Actions build matrix.
 #
 # Usage:
-#   docker build -f Dockerfile.ci -t fast-mempalace-ci .
-#   docker run --rm fast-mempalace-ci
+#   docker build -f Dockerfile.ci -t memxt-ci .
+#   docker run --rm memxt-ci
 #
 # Or via the helper:
 #   ./scripts/ci-local.sh
@@ -89,4 +89,4 @@ RUN find lib/llama.cpp/build -name '*.a' -print | head -20 \
 
 RUN zig build --release=fast
 
-CMD ["sh", "-c", "./zig-out/bin/fast-mempalace init && ./zig-out/bin/fast-mempalace stats"]
+CMD ["sh", "-c", "./zig-out/bin/memxt init && ./zig-out/bin/memxt stats"]

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2026 debpalash
+Copyright (c) 2026 Yupcha
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,17 +1,17 @@
 <div align="center">
-  <img src="assets/logo.svg" alt="Fast MemPalace Logo" width="350"/>
-  <h1>fast-mempalace</h1>
+  <img src="assets/logo.svg" alt="memxt Logo" width="350"/>
+  <h1>memxt</h1>
   <p><b>Local-first long-term memory for AI coding agents.</b><br/>
   Your agent remembers your codebase and decisions across sessions — in a single static binary. No cloud, no Python, nothing leaves your machine.</p>
   <br/>
-  <img src="assets/demo.gif" alt="fast-mempalace semantic recall demo" width="760"/>
+  <img src="assets/demo.gif" alt="memxt semantic recall demo" width="760"/>
 </div>
 
 ---
 
 Most AI coding sessions start amnesiac: you re-explain the architecture, re-state why
 you chose SQLite over Postgres, and the agent still contradicts last week's decision.
-**fast-mempalace** gives the agent a persistent, on-device memory it can search and
+**memxt** gives the agent a persistent, on-device memory it can search and
 write to — wired into Claude Code via MCP tools and session hooks.
 
 - 🧠 **Remembers across sessions** — semantic recall over everything you've mined or saved.
@@ -22,20 +22,20 @@ write to — wired into Claude Code via MCP tools and session hooks.
 ## ⚡ Install
 
 ```bash
-curl -fsSL https://raw.githubusercontent.com/debpalash/fast-mempalace/main/install.sh | bash
+curl -fsSL https://raw.githubusercontent.com/Yupcha/memxt/main/install.sh | bash
 ```
 
 Detects `darwin|linux` × `x86_64|aarch64`. Binary + embedding model land in
-`~/.fast-mempalace/`.
+`~/.memxt/`.
 
 ## 🤖 Use with Claude Code (the main event)
 
 Install the plugin — it wires up the MCP memory tools, the session hooks, a skill, and
 slash commands:
 
 ```text
-/plugin marketplace add debpalash/fast-mempalace
-/plugin install fast-mempalace
+/plugin marketplace add Yupcha/memxt
+/plugin install memxt
 ```
 
 You now have:
@@ -52,11 +52,11 @@ You now have:
 Optionally seed memory from a codebase:
 
 ```bash
-~/.fast-mempalace/bin/fast-mempalace mine . my-project
+~/.memxt/bin/memxt mine . my-project
 ```
 
 > Works with any MCP client (Cursor, Zed, Windsurf, …) too — point it at
-> `fast-mempalace mcp`.
+> `memxt mcp`.
 
 ## 🧩 How it works
 
@@ -89,30 +89,30 @@ Apple Silicon · Metal · cold process unless noted. Methodology →
 ## 📦 CLI
 
 ```text
-fast-mempalace init                  Initialize the palace database
-fast-mempalace mine <path> [wing]    Mine a file or directory into the palace
-fast-mempalace search <query>        Semantic search
-fast-mempalace wake-up [--wing X]    Print the wake-up context
-fast-mempalace stats                 Palace statistics
-fast-mempalace mcp                   Start the MCP server (stdio JSON-RPC)
-fast-mempalace hook                  Run a Claude Code hook (JSON stdin/stdout)
-fast-mempalace kg [subject]          Query the knowledge graph
+memxt init                  Initialize the palace database
+memxt mine <path> [wing]    Mine a file or directory into the palace
+memxt search <query>        Semantic search
+memxt wake-up [--wing X]    Print the wake-up context
+memxt stats                 Palace statistics
+memxt mcp                   Start the MCP server (stdio JSON-RPC)
+memxt hook                  Run a Claude Code hook (JSON stdin/stdout)
+memxt kg [subject]          Query the knowledge graph
 ```
 
 ## ⚙️ Configuration
 
-Reads `fast-mempalace.yaml` (falls back to `mempalace.yaml`). Environment variables
+Reads `memxt.yaml` (falls back to `mempalace.yaml`). Environment variables
 override everything — this is how the plugin pins one global palace regardless of the
 project directory:
 
 ```bash
-FAST_MEMPALACE_DB=~/.fast-mempalace/palace.db           # database path
-FAST_MEMPALACE_MODEL=~/.fast-mempalace/lib/minilm.gguf  # 384-dim GGUF embedder
-FAST_MEMPALACE_WING=my-project                          # default wing
+MEMXT_DB=~/.memxt/palace.db           # database path
+MEMXT_MODEL=~/.memxt/lib/minilm.gguf  # 384-dim GGUF embedder
+MEMXT_WING=my-project                          # default wing
 ```
 
 ```yaml
-database_path: "fast-mempalace.db"
+database_path: "memxt.db"
 model_path: "lib/minilm.gguf"
 default_wing: "production"
 ```
@@ -125,8 +125,8 @@ The embedding model must be **384-dim** (MiniLM-L6-v2); the vector table is decl
 Needs `zig 0.16.0` + `cmake`.
 
 ```bash
-git clone --recursive https://github.com/debpalash/fast-mempalace
-cd fast-mempalace
+git clone --recursive https://github.com/Yupcha/memxt
+cd memxt
 
 # 1) Build the statically-linked llama.cpp backend (once)
 cmake -S lib/llama.cpp -B lib/llama.cpp/build \
@@ -141,7 +141,7 @@ mkdir -p lib && curl -L -o lib/minilm.gguf \
 
 # 3) Build
 zig build --release=fast
-./zig-out/bin/fast-mempalace stats
+./zig-out/bin/memxt stats
 ```
 
 (On Linux, use `-DGGML_METAL=OFF -DGGML_BLAS=OFF`.)

ROADMAP.md

@@ -1,6 +1,6 @@
 # 🗺️ Roadmap — Parity & Outmatch
 
-Path from `fast-mempalace` today to **full drop-in parity** with the upstream Python `mempalace` package, then to **outmatching** it with capabilities the Python stack cannot reach.
+Path from `memxt` today to **full drop-in parity** with the upstream Python `mempalace` package, then to **outmatching** it with capabilities the Python stack cannot reach.
 
 Legend: `[x]` done · `[~]` partial · `[ ]` planned
 
@@ -19,10 +19,10 @@ against Claude Code 2.1.191.
 - [x] `wake-up [--wing X]` — L0+L1 context loader (~600–900 tok)
 - [x] **`mcp` — real MCP server**: `memory_search` / `memory_store` / `memory_wake_up` / `memory_stats`, lazy model load, protocol-version echo (was a hardcoded stub)
 - [x] **`hook` — real Claude Code protocol**: SessionStart injects wake-up via `additionalContext`; PreCompact reads the transcript and **auto-saves** the tail (was a custom protocol + a nag)
-- [x] **Claude Code plugin** — `claude-plugin/` + marketplace manifest: MCP server + hooks + `using-memory` skill + `/remember` `/recall` commands; one global palace via `FAST_MEMPALACE_DB`/`_MODEL` env overrides
+- [x] **Claude Code plugin** — `claude-plugin/` + marketplace manifest: MCP server + hooks + `using-memory` skill + `/remember` `/recall` commands; one global palace via `MEMXT_DB`/`_MODEL` env overrides
 - [x] `kg [subject]` — knowledge-graph relationship query (manual population)
 - [x] `instructions` — memory-instruction emitter
-- [x] `mempalace.yaml` + `fast-mempalace.yaml` config + env-var overrides
+- [x] `mempalace.yaml` + `memxt.yaml` config + env-var overrides
 - [x] MIT license, GitHub Actions CI, one-line curl installer (now fetches a **384-dim** model)
 - [x] Honest benchmarks vs the real engine (`BENCHMARK.md`); retrieval 7/7 top-1 on a paraphrase test
 
@@ -40,13 +40,13 @@ Close every remaining gap with the upstream `pip install mempalace` surface. Eac
 - [ ] **`init` vs `stats`** — upstream uses `init`; alias our `stats` where appropriate
 - [ ] **Incremental re-mining** — content-hash skip for unchanged files (currently re-embeds all)
 - [ ] **`forget <id|wing>`** — evict drawers
-- [ ] **Export / import** — `fast-mempalace export <path>` JSONL dump + re-ingest round-trip
+- [ ] **Export / import** — `memxt export <path>` JSONL dump + re-ingest round-trip
 - [ ] **Ignore-pattern parity** — `.gitignore`-style globs matching upstream semantics
 - [ ] **Config schema audit** — every upstream yaml key respected or rejected with a diagnostic
 - [ ] **Python-parity output strings** — exit codes, stderr format, progress-bar layout for script consumers
 - [ ] **Embedding model swap** — allow upstream's default model name via `model: <name>` resolving to HF URL
 
-**Definition of done:** a user can `pip uninstall mempalace && curl ... | bash && ln -s .../fast-mempalace .../mempalace` and every script in their pipeline keeps working unchanged.
+**Definition of done:** a user can `pip uninstall mempalace && curl ... | bash && ln -s .../memxt .../mempalace` and every script in their pipeline keeps working unchanged.
 
 ---
 
@@ -59,23 +59,23 @@ Ship features upstream Python cannot match without rewriting. Each lands a capab
 - [ ] **Batched embedding kernel** — vectorize mine across N files per GPU call (target: 10× mine throughput vs current 200×)
 - [ ] **Incremental vector index** — sqlite-vec HNSW params tuned per drawer-count bucket
 - [ ] **Zero-copy mmap ingest** — large file mining without full read-into-RAM
-- [ ] **Compile-time schema** — Zig comptime validation of `fast-mempalace.yaml`; bad config fails at build, not runtime
+- [ ] **Compile-time schema** — Zig comptime validation of `memxt.yaml`; bad config fails at build, not runtime
 
 ### v0.4 — Reach Beyond CLI
 
-- [ ] **Watch mode** — `fast-mempalace watch <path>` file-system events → auto re-mine (upstream Python blocks on ChromaDB lock; we don't)
-- [ ] **Embedded HTTP API** — `fast-mempalace serve --port 8080` pure Zig handler, <5 MB RAM overhead
+- [ ] **Watch mode** — `memxt watch <path>` file-system events → auto re-mine (upstream Python blocks on ChromaDB lock; we don't)
+- [ ] **Embedded HTTP API** — `memxt serve --port 8080` pure Zig handler, <5 MB RAM overhead
 - [ ] **Web UI** — single-file static dashboard shipped inside binary (SQLite browser + search box)
 - [ ] **Hybrid search** — BM25 + vector fusion (upstream is vector-only)
 - [ ] **Time-scoped queries** — `--since 2026-01-01`, `--until`, decay-weighted ranking
 
 ### v0.5 — Ecosystem & Distribution
 
-- [ ] **Homebrew formula** — `brew install fast-mempalace`
+- [ ] **Homebrew formula** — `brew install memxt`
 - [ ] **Docker image** — ~15 MB distroless image (vs upstream ~1.2 GB Python+ML)
 - [ ] **Shell completions** — zsh / bash / fish
 - [x] **Claude Code plugin** — `claude-plugin/` wires MCP + hooks + skill + slash commands (v0.2)
-- [ ] **Plugin SDK** — stable `lib/fast_mempalace.h` C ABI for 3rd-party languages
+- [ ] **Plugin SDK** — stable `lib/memxt.h` C ABI for 3rd-party languages
 
 ### v0.6+ — Intelligence Layer
 

build.zig

@@ -8,7 +8,7 @@ pub fn build(b: *std.Build) void {
 
     // ── Main executable ──
     const exe = b.addExecutable(.{
-        .name = "fast-mempalace",
+        .name = "memxt",
         .root_module = b.createModule(.{
             .root_source_file = b.path("src/main.zig"),
             .target = target,
@@ -25,7 +25,7 @@ pub fn build(b: *std.Build) void {
     run_cmd.step.dependOn(b.getInstallStep());
     if (b.args) |args| run_cmd.addArgs(args);
 
-    const run_step = b.step("run", "Run fast-mempalace");
+    const run_step = b.step("run", "Run memxt");
     run_step.dependOn(&run_cmd.step);
 
     // ── Tests ──

build.zig.zon

@@ -1,7 +1,7 @@
 .{
-    .name = .fast_mempalace,
+    .name = .memxt,
     .version = "0.1.0",
-    .fingerprint = 0x182ae31931ffe1cb,
+    .fingerprint = 0x5bf95d78bca15787,
     .minimum_zig_version = "0.16.0",
     .paths = .{
         "build.zig",

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

@@ -1,12 +1,12 @@
 {
   "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
-  "name": "fast-mempalace",
-  "displayName": "Fast MemPalace — Local Memory",
+  "name": "memxt",
+  "displayName": "memxt — Local Memory",
   "version": "0.2.0",
   "description": "Persistent, private, local-first memory for Claude Code. Your agent remembers your code and decisions across sessions — one static binary, no cloud, nothing leaves your machine.",
-  "author": { "name": "MemPalace", "url": "https://github.com/debpalash" },
-  "homepage": "https://github.com/debpalash/fast-mempalace",
-  "repository": "https://github.com/debpalash/fast-mempalace",
+  "author": { "name": "Yupcha", "url": "https://github.com/Yupcha" },
+  "homepage": "https://github.com/Yupcha/memxt",
+  "repository": "https://github.com/Yupcha/memxt",
   "license": "MIT",
   "keywords": ["memory", "mcp", "local-first", "claude-code", "embeddings", "privacy"],
   "mcpServers": "./.mcp.json",