qmdr 1.0.0

package/.claude-plugin/marketplace.json

@@ -0,0 +1,29 @@
+{
+	"name": "qmdr",
+	"owner": {
+		"name": "uf-hy",
+		"email": "uf-hy@users.noreply.github.com"
+	},
+	"plugins": [
+		{
+			"name": "qmdr",
+			"source": "./",
+			"description": "Search personal markdown files with hybrid BM25 + vector + cloud LLM reranking.",
+			"version": "1.0.0",
+			"author": {
+				"name": "uf-hy",
+				"email": "uf-hy@users.noreply.github.com"
+			},
+			"repository": "https://github.com/uf-hy/qmdr",
+			"license": "MIT",
+			"keywords": ["markdown", "search", "qmd", "memory", "reranking", "remote"],
+			"skills": ["./skills/"],
+			"mcpServers": {
+				"qmd": {
+					"command": "qmd",
+					"args": ["mcp"]
+				}
+			}
+		}
+	]
+}

package/.env.example

@@ -0,0 +1,85 @@
+# QMDR Environment Variables
+# Copy to .env, fill in your API keys, then `source .env`
+#
+# 💡 Three config modes (pick one):
+#
+# Mode A: SiliconFlow only (simplest, one key)
+#   → Embedding / Query Expansion / Rerank all via SiliconFlow
+#
+# Mode B: SiliconFlow + Gemini (recommended)
+#   → Embedding / Expansion via SiliconFlow, Rerank via Gemini
+#
+# Mode C: OpenAI-compatible API only
+#   → OpenAI, Azure, Volcengine, any relay proxy
+#
+# 💡 三种配置方式（任选其一）：
+#
+# 方式一：只用 SiliconFlow（最简，一个 Key）
+# 方式二：SiliconFlow + Gemini（推荐，各取所长）
+# 方式三：只用 OpenAI 兼容 API（火山引擎、中转站等）
+
+# ============================================================
+# SiliconFlow
+# China: https://cloud.siliconflow.cn (free tier)
+# International: https://cloud.siliconflow.com (free tier)
+# ============================================================
+
+QMD_SILICONFLOW_API_KEY=sk-your-siliconflow-key-here
+
+# International users: uncomment to use the global endpoint
+# QMD_SILICONFLOW_BASE_URL=https://api.siliconflow.com/v1
+
+# ============================================================
+# Gemini (https://ai.google.dev — free tier: 15 RPM / 1M tokens/day)
+# ============================================================
+
+# QMD_GEMINI_API_KEY=your-gemini-key-here
+# QMD_GEMINI_MODEL=gemini-2.5-flash
+# QMD_GEMINI_BASE_URL=https://generativelanguage.googleapis.com
+
+# ============================================================
+# OpenAI-compatible API
+# Supports: OpenAI, Azure, Volcengine, relay proxies, etc.
+# ============================================================
+
+# QMD_OPENAI_API_KEY=your-key-here
+# QMD_OPENAI_BASE_URL=https://api.openai.com/v1
+# QMD_OPENAI_MODEL=gpt-4o-mini
+# QMD_OPENAI_EMBED_MODEL=text-embedding-3-small
+
+# ============================================================
+# Provider routing (auto-detected from keys, usually not needed)
+# ============================================================
+
+# QMD_EMBED_PROVIDER=siliconflow          # siliconflow / openai
+# QMD_QUERY_EXPANSION_PROVIDER=siliconflow # siliconflow / gemini / openai
+# QMD_RERANK_PROVIDER=gemini               # siliconflow / gemini / openai
+# QMD_RERANK_MODE=llm                      # llm (chat model) / rerank (dedicated rerank API)
+
+# ============================================================
+# Model overrides (sensible defaults provided)
+# ============================================================
+
+# SiliconFlow models
+# QMD_SILICONFLOW_EMBED_MODEL=Qwen/Qwen3-Embedding-8B
+# QMD_SILICONFLOW_QUERY_EXPANSION_MODEL=zai-org/GLM-4.5-Air
+# QMD_SILICONFLOW_RERANK_MODEL=BAAI/bge-reranker-v2-m3   # when QMD_RERANK_MODE=rerank
+# QMD_LLM_RERANK_MODEL=zai-org/GLM-4.5-Air               # when QMD_RERANK_MODE=llm
+
+# ============================================================
+# Performance tuning (sensible defaults provided)
+# ============================================================
+
+# QMD_CHUNK_SIZE_TOKENS=200           # Tokens per chunk (default: 200)
+# QMD_CHUNK_OVERLAP_TOKENS=40         # Overlap between chunks (default: 40)
+# QMD_EMBED_BATCH_SIZE=32             # Embedding batch size (default: 32)
+# QMD_RERANK_DOC_LIMIT=40             # Max docs for reranking (default: 40)
+# QMD_RERANK_CHUNKS_PER_DOC=3         # Chunks per doc for reranking (default: 3)
+
+# ============================================================
+# Paths (auto-detected, override if needed)
+# ============================================================
+
+# QMD_CONFIG_DIR=~/.config/qmd        # Config directory (index.yml location)
+# XDG_CACHE_HOME=~/.cache             # Cache directory (database at $XDG_CACHE_HOME/qmd/index.sqlite)
+# QMD_SQLITE_VEC_PATH=                # sqlite-vec native extension path (auto-detected)

package/.gitattributes

@@ -0,0 +1,3 @@
+
+# Use bd merge for beads JSONL files
+.beads/issues.jsonl merge=beads

package/.github/workflows/release.yml

@@ -0,0 +1,77 @@
+name: Release
+
+on:
+  push:
+    tags:
+      - 'v*'
+
+permissions:
+  contents: write
+
+jobs:
+  build:
+    strategy:
+      matrix:
+        include:
+          - os: macos-latest
+            target: darwin-arm64
+            artifact: qmd-darwin-arm64
+          - os: macos-13
+            target: darwin-x64
+            artifact: qmd-darwin-x64
+          - os: ubuntu-latest
+            target: linux-x64
+            artifact: qmd-linux-x64
+          - os: ubuntu-latest
+            target: linux-arm64
+            artifact: qmd-linux-arm64
+
+    runs-on: ${{ matrix.os }}
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: oven-sh/setup-bun@v2
+        with:
+          bun-version: latest
+
+      - name: Install dependencies
+        run: bun install
+
+      - name: Build binary
+        run: |
+          if [ "${{ matrix.target }}" = "linux-arm64" ]; then
+            bun build --compile --target=bun-linux-arm64 src/qmd.ts --outfile ${{ matrix.artifact }}
+          else
+            bun build --compile src/qmd.ts --outfile ${{ matrix.artifact }}
+          fi
+
+      - name: Test binary
+        if: matrix.target != 'linux-arm64'
+        run: |
+          ./${{ matrix.artifact }} --help
+          ./${{ matrix.artifact }} status 2>/dev/null || true
+          echo "✅ Binary runs successfully"
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: ${{ matrix.artifact }}
+          path: ${{ matrix.artifact }}
+
+  release:
+    needs: build
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          merge-multiple: true
+
+      - name: Make binaries executable
+        run: chmod +x qmd-*
+
+      - name: Create Release
+        uses: softprops/action-gh-release@v2
+        with:
+          files: qmd-*
+          generate_release_notes: true
+          draft: false

package/AI-SETUP.md

@@ -0,0 +1,466 @@
+# QMDR - AI Setup Guide
+
+> **For AI coding agents.** Humans: just paste this file's URL to your AI and let it handle the rest.
+
+QMDR is a fork of [tobi/qmd](https://github.com/tobi/qmd) — a hybrid search engine for markdown files (BM25 + vector + LLM reranking). QMDR replaces local GGUF models with cloud APIs, so there's no GPU needed and reranking quality is much higher.
+
+---
+
+## Phase 0: Detect Environment
+
+Before starting, determine the following:
+
+### 1. Which AI client is the user using?
+
+| Client | Integration | Config file for usage tips |
+|--------|------------|---------------------------|
+| **OpenClaw** | Native memory backend | `TOOLS.md` in workspace |
+| **Claude Code** | Skill (SKILL.md) | `CLAUDE.md` in project root |
+| **OpenCode** | Skill (SKILL.md) | `AGENTS.md` in project root |
+| **Other** | CLI only | N/A |
+
+### 2. What region is the user in?
+
+Detect via system locale, timezone, or ask directly. This affects:
+- **China** → Use `https://cloud.siliconflow.cn` for registration, `https://api.siliconflow.cn/v1` as API endpoint. Bun install may need mirror.
+- **International** → Use `https://cloud.siliconflow.com` for registration, `https://api.siliconflow.com/v1` as API endpoint.
+
+### 3. Interaction style
+
+| Client | How to present choices |
+|--------|----------------------|
+| **OpenClaw** | Use inline buttons (`buttons` parameter in message tool) |
+| **Claude Code / OpenCode** | Numbered list — ask user to reply with number |
+| **Other** | Y/N or numbered input |
+
+---
+
+## Phase 1: Quick Install or Step-by-Step?
+
+Ask the user:
+
+> **Do you want one-click install or step-by-step setup?**
+>
+> **One-click:** Provide your API keys and I'll handle everything — install Bun (if missing), install QMDR, configure all providers, verify with doctor.
+>
+> **Step-by-step:** I'll walk you through each decision.
+
+### One-click install flow
+
+Collect upfront:
+1. SiliconFlow API key (required for embedding)
+2. Gemini API key (recommended for reranking) — or skip to use SiliconFlow for everything
+3. Any custom OpenAI-compatible endpoint (optional)
+
+Then execute Phase 2 → 3 → 4 → 5 → 6 → 7 automatically, using defaults for all choices.
+
+### Step-by-step flow
+
+Continue to Phase 2.
+
+---
+
+## Phase 2: Install QMDR
+
+### Check for Bun
+
+```bash
+bun --version 2>/dev/null || echo "NOT_INSTALLED"
+```
+
+**If Bun is installed:** skip to "Install QMDR" below.
+
+**If Bun is NOT installed:** ask the user:
+
+> Bun is required to run QMDR. Install it?
+>
+> 1. **Yes, install Bun** (recommended)
+> 2. **No, download pre-built binary instead**
+
+#### Install Bun
+
+```bash
+# International
+curl -fsSL https://bun.sh/install | bash
+
+# China (if above fails or is slow)
+curl -fsSL https://bun.sh/install | BUN_INSTALL_MIRROR=https://npmmirror.com/mirrors/bun bash
+```
+
+After install, verify: `bun --version`
+
+#### Install QMDR (two paths)
+
+**Path A: Via Bun (recommended)**
+```bash
+bun install -g github:uf-hy/qmdr
+qmd --help  # verify
+```
+
+**Path B: Pre-built binary (no Bun needed)**
+```bash
+# macOS arm64
+curl -L https://github.com/uf-hy/qmdr/releases/latest/download/qmd-darwin-arm64 -o /usr/local/bin/qmd
+chmod +x /usr/local/bin/qmd
+
+# Linux x64
+curl -L https://github.com/uf-hy/qmdr/releases/latest/download/qmd-linux-x64 -o /usr/local/bin/qmd
+chmod +x /usr/local/bin/qmd
+```
+
+⚠️ Binary users: `sqlite-vec` native extension must be discoverable. macOS: `brew install sqlite`. Or set `QMD_SQLITE_VEC_PATH=/path/to/vec0.dylib`.
+
+---
+
+## Phase 3: Configure Embedding (SiliconFlow)
+
+SiliconFlow provides the embedding model. Ask:
+
+> **Do you have a SiliconFlow API key?**
+>
+> 1. **Yes** — I'll enter my key
+> 2. **No** — show me how to get one
+> 3. **Custom provider** — I want to use a different OpenAI-compatible endpoint
+
+### Option 1: User has key
+
+```bash
+export QMD_SILICONFLOW_API_KEY=sk-user-provided-key
+```
+
+### Option 2: Guide to get key
+
+**China users:**
+> Register at https://cloud.siliconflow.cn (no affiliate link, the author was too lazy to set one up)
+> 1. Sign up with phone number or email
+> 2. Go to API Keys page → Create new key
+> 3. Free tier includes embedding models
+
+**International users:**
+> Register at https://cloud.siliconflow.com (no affiliate link)
+> 1. Sign up with email
+> 2. Go to API Keys page → Create new key
+> 3. Free tier includes embedding models
+
+After user provides key:
+```bash
+export QMD_SILICONFLOW_API_KEY=sk-user-provided-key
+```
+
+**International users — additional API endpoint config:**
+```bash
+export QMD_SILICONFLOW_BASE_URL=https://api.siliconflow.com/v1
+```
+(China users don't need this — the default `https://api.siliconflow.cn/v1` works)
+
+### Option 3: Custom provider
+
+Ask for:
+1. API endpoint URL (must be OpenAI-compatible)
+2. API key
+
+```bash
+export QMD_OPENAI_API_KEY=user-key
+export QMD_OPENAI_BASE_URL=https://their-endpoint.com/v1
+export QMD_EMBED_PROVIDER=openai
+```
+
+Test connectivity:
+```bash
+curl -s "$QMD_OPENAI_BASE_URL/models" -H "Authorization: Bearer $QMD_OPENAI_API_KEY" | head -5
+```
+
+### Configure embedding model
+
+Default: `Qwen/Qwen3-Embedding-8B` (on SiliconFlow, free)
+
+Ask:
+> Keep the default embedding model (`Qwen/Qwen3-Embedding-8B`), or choose your own?
+>
+> If you don't know which to pick, the default is good.
+
+If user wants to change:
+```bash
+export QMD_SILICONFLOW_EMBED_MODEL=their-chosen-model
+# or for custom provider:
+export QMD_OPENAI_EMBED_MODEL=their-chosen-model
+```
+
+### Configure embedding dimensions
+
+> Do you know the output dimensions of your embedding model?
+>
+> 1. **Use default** (auto-detect from model — recommended)
+> 2. **I know the dimensions** — let me enter it
+> 3. **I don't know** — please look it up for me
+
+If option 3: search the web for "{model_name} embedding dimensions" and inform the user.
+
+Note: If dimensions change after initial indexing, user must run `qmd embed -f` to rebuild.
+
+### Configure chunk size
+
+Default: `200` tokens per chunk, `40` tokens overlap.
+
+> Chunk size controls how documents are split for embedding.
+>
+> - **Default: 200 tokens** (recommended for most use cases)
+> - Larger chunks = more context per result, but less precise
+> - Smaller chunks = more precise, but may lose context
+
+```bash
+# Only if user wants to change defaults:
+export QMD_CHUNK_SIZE_TOKENS=200      # default
+export QMD_CHUNK_OVERLAP_TOKENS=40    # default
+```
+
+---
+
+## Phase 4: Configure Query Expansion
+
+Query expansion rewrites the user's search query into multiple variations (keyword + semantic + hypothetical) to improve recall.
+
+Default: `GLM-4.5-Air` on SiliconFlow (~¥1/month, fast, good quality).
+
+Ask:
+> Use the default query expansion model (`GLM-4.5-Air` on SiliconFlow), or use your own?
+>
+> 1. **Default** (GLM-4.5-Air on SiliconFlow — recommended)
+> 2. **Use the reranking provider's model** (configured in next step)
+> 3. **Custom** — I want to specify a model
+
+If option 2: will be configured after Phase 5.
+
+If option 3:
+```bash
+export QMD_QUERY_EXPANSION_PROVIDER=openai  # or gemini
+# For OpenAI-compatible:
+export QMD_OPENAI_MODEL=their-model-name
+# For Gemini:
+export QMD_GEMINI_API_KEY=their-key
+```
+
+---
+
+## Phase 5: Configure Reranking
+
+Reranking uses a large language model to judge which search results are truly relevant. This is where QMDR's quality advantage comes from — cloud LLMs are much better at this than the original QMD's 0.6B local model.
+
+**Default: Gemini 2.5 Flash** with thinking budget set to 0 (fast, high quality, generous free tier).
+
+> **Do you have a Gemini API key?**
+>
+> 1. **Yes** — I'll enter my key
+> 2. **No** — show me how to get one
+> 3. **Skip Gemini** — use SiliconFlow or my own model for reranking
+
+### Option 1: User has Gemini key
+
+```bash
+export QMD_GEMINI_API_KEY=user-key
+export QMD_RERANK_PROVIDER=gemini
+```
+
+### Option 2: Guide to get Gemini key
+
+> Get a free Gemini API key at https://ai.google.dev
+> 1. Sign in with Google account
+> 2. Click "Get API key" → Create key
+> 3. Free tier: 15 RPM / 1M tokens per day (more than enough for reranking)
+
+Note for China users: Gemini API may require a proxy. If the user has a proxy:
+```bash
+export QMD_GEMINI_BASE_URL=https://their-proxy-endpoint
+```
+
+If no proxy available, recommend Option 3 instead.
+
+### Option 3: Alternative reranking
+
+**Using SiliconFlow LLM rerank (no extra key needed):**
+```bash
+export QMD_RERANK_PROVIDER=siliconflow
+export QMD_RERANK_MODE=llm
+```
+
+**Using a dedicated rerank model API (e.g. BAAI/bge-reranker):**
+```bash
+export QMD_RERANK_PROVIDER=siliconflow
+export QMD_RERANK_MODE=rerank
+export QMD_SILICONFLOW_RERANK_MODEL=BAAI/bge-reranker-v2-m3
+```
+
+**Using OpenAI-compatible endpoint:**
+```bash
+export QMD_RERANK_PROVIDER=openai
+export QMD_RERANK_MODE=llm
+export QMD_OPENAI_API_KEY=their-key
+export QMD_OPENAI_BASE_URL=https://their-endpoint/v1
+```
+
+For Claude Code / OpenCode users: you can reuse whichever model API you're already paying for.
+
+### Model selection
+
+Default reranking model: `gemini-2.5-flash` with `thinkingBudget: 0` (no reasoning overhead, pure relevance judgment).
+
+> Keep the default reranking model (`gemini-2.5-flash`), or change it?
+
+If user wants to change:
+```bash
+export QMD_GEMINI_MODEL=their-model      # for Gemini provider
+export QMD_LLM_RERANK_MODEL=their-model  # for SiliconFlow/OpenAI LLM rerank
+```
+
+---
+
+## Phase 6: Doctor Verification
+
+Run the diagnostic:
+
+```bash
+qmd doctor
+```
+
+**Read the output carefully.** Check for:
+
+| Status | Meaning | Action |
+|--------|---------|--------|
+| ✅ | Provider working | Good |
+| ❌ | Provider failed | Fix config or key |
+| Speed > 5s | Model is slow | Suggest switching to a faster/non-thinking model |
+
+If any provider shows errors, go back to the relevant Phase and fix.
+
+If speed > 5s for reranking or query expansion:
+> Your {component} is taking {time}s — this may slow down searches. Consider switching to a faster model (non-thinking models like `gemini-2.5-flash` with thinkingBudget=0 are recommended).
+
+Optional quality check:
+```bash
+qmd doctor --bench
+```
+This runs test queries and evaluates expansion quality.
+
+---
+
+## Phase 7: Client Integration
+
+### OpenClaw
+
+QMDR is a **native memory backend** for OpenClaw — no MCP needed. OpenClaw spawns QMDR directly.
+
+**→ See [docs/setup-openclaw.md](docs/setup-openclaw.md) for detailed configuration.**
+
+After setup, add this to `TOOLS.md` in the workspace:
+
+```markdown
+## Memory Search (QMDR)
+
+- **Always use natural language** for memory_search queries — not keyword concatenation
+- ✅ Good: "what did we discuss about the server migration last week"
+- ❌ Bad: "server migration discussion last week"
+- The hybrid pipeline (query expansion + vector + rerank) is optimized for natural language
+- `memory_search` is an internal OpenClaw tool, never a shell command
+```
+
+### Claude Code
+
+Install the QMDR skill:
+
+```bash
+mkdir -p ~/.claude/skills/qmdr
+curl -sL https://raw.githubusercontent.com/uf-hy/qmdr/main/skills/qmdr/SKILL.md \
+  -o ~/.claude/skills/qmdr/SKILL.md
+```
+
+After setup, add this to `CLAUDE.md` in the project root:
+
+```markdown
+## Memory Search (QMDR)
+
+- Use `qmd query "natural language question"` for best results
+- Always use natural language — not keyword concatenation
+- ✅ Good: `qmd query "what authentication method did we decide on"`
+- ❌ Bad: `qmd query "authentication method decision"`
+- Use `qmd search` for fast keyword-only lookups
+- Use `qmd get "#docid"` to retrieve full documents from search results
+```
+
+### OpenCode
+
+Install the QMDR skill:
+
+```bash
+mkdir -p ~/.config/opencode/skills/qmdr
+curl -sL https://raw.githubusercontent.com/uf-hy/qmdr/main/skills/qmdr/SKILL.md \
+  -o ~/.config/opencode/skills/qmdr/SKILL.md
+```
+
+After setup, add this to `AGENTS.md` in the project root:
+
+```markdown
+## Memory Search (QMDR)
+
+- Use `qmd query "natural language question"` for best results
+- Always use natural language — not keyword concatenation
+- ✅ Good: `qmd query "what authentication method did we decide on"`
+- ❌ Bad: `qmd query "authentication method decision"`
+- Use `qmd search` for fast keyword-only lookups
+- Use `qmd get "#docid"` to retrieve full documents from search results
+```
+
+---
+
+## Environment Variables Reference
+
+| Variable | Default | Purpose |
+|----------|---------|---------||
+| **API Keys** | | |
+| `QMD_SILICONFLOW_API_KEY` | — | SiliconFlow |
+| `QMD_GEMINI_API_KEY` | — | Google Gemini |
+| `QMD_OPENAI_API_KEY` | — | OpenAI-compatible |
+| **Provider routing** | | |
+| `QMD_EMBED_PROVIDER` | auto | `siliconflow` / `openai` |
+| `QMD_QUERY_EXPANSION_PROVIDER` | auto | `siliconflow` / `gemini` / `openai` |
+| `QMD_RERANK_PROVIDER` | auto | `siliconflow` / `gemini` / `openai` |
+| `QMD_RERANK_MODE` | `llm` | `llm` (chat model) / `rerank` (dedicated rerank API) |
+| **Custom endpoints** | | |
+| `QMD_SILICONFLOW_BASE_URL` | `https://api.siliconflow.cn/v1` | SiliconFlow endpoint (international: use `api.siliconflow.com`) |
+| `QMD_GEMINI_BASE_URL` | Google default | Custom Gemini endpoint |
+| `QMD_OPENAI_BASE_URL` | `api.openai.com/v1` | Custom OpenAI endpoint |
+| **Model overrides** | | |
+| `QMD_SILICONFLOW_EMBED_MODEL` | `Qwen/Qwen3-Embedding-8B` | Embedding model |
+| `QMD_SILICONFLOW_QUERY_EXPANSION_MODEL` | `zai-org/GLM-4.5-Air` | Query expansion |
+| `QMD_SILICONFLOW_RERANK_MODEL` | `BAAI/bge-reranker-v2-m3` | Dedicated rerank model |
+| `QMD_LLM_RERANK_MODEL` | `zai-org/GLM-4.5-Air` | LLM rerank model |
+| `QMD_GEMINI_MODEL` | `gemini-2.5-flash` | Gemini model (thinkingBudget=0 by default) |
+| `QMD_OPENAI_MODEL` | `gpt-4o-mini` | OpenAI chat model |
+| `QMD_OPENAI_EMBED_MODEL` | `text-embedding-3-small` | OpenAI embed model |
+| **Tuning** | | |
+| `QMD_CHUNK_SIZE_TOKENS` | `200` | Tokens per chunk |
+| `QMD_CHUNK_OVERLAP_TOKENS` | `40` | Overlap between chunks |
+| `QMD_EMBED_BATCH_SIZE` | `32` | Embedding batch size |
+| `QMD_RERANK_DOC_LIMIT` | `40` | Max docs for reranking |
+| `QMD_RERANK_CHUNKS_PER_DOC` | `3` | Chunks per doc for reranking |
+| **Paths** | | |
+| `QMD_SQLITE_VEC_PATH` | auto | sqlite-vec native extension path |
+
+**If you change the embedding model, run `qmd embed -f` to rebuild the vector index.**
+
+## Troubleshooting
+
+| Problem | Fix |
+|---------|-----|
+| `qmd: command not found` | `bun link`, ensure `~/.bun/bin` in PATH |
+| sqlite-vec load error | macOS: `brew install sqlite`; or set `QMD_SQLITE_VEC_PATH` |
+| sqlite-vec `/$bunfs/root/` error | Switch to source mode (bun install) instead of compiled binary |
+| Dimension mismatch | `qmd embed -f` to rebuild with current model |
+| Slow queries (>5s) | Run `qmd doctor` — switch to non-thinking models |
+| `Script not found "query"` | Don't use `"bun /path/to/qmd.ts"` as command; use direct path with shebang |
+
+## Links
+
+- Original project: https://github.com/tobi/qmd
+- SiliconFlow (China): https://cloud.siliconflow.cn
+- SiliconFlow (International): https://cloud.siliconflow.com
+- Gemini API: https://ai.google.dev

package/LICENSE

@@ -0,0 +1,22 @@
+MIT License
+
+Copyright (c) 2025 QMDR Contributors
+Based on QMD by Tobias Lütke (https://github.com/tobi/qmd)
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.