autosnippet 3.3.9 → 3.4.1

package/README.md

@@ -14,7 +14,7 @@ Extract patterns from your codebase into a knowledge base that AI coding assista
 
 ---
 
-- [Why](#why) · [Getting Started](#getting-started) · [Using in IDE](#using-in-ide) · [Evolution Architecture](#evolution-architecture) · [Engineering Capabilities](#engineering-capabilities) · [IDE Support](#ide-support) · [Docs](#docs)
+- [Why](#why) · [Getting Started](#getting-started) · [Using in IDE](#using-in-ide) · [Evolution Architecture](#evolution-architecture) · [Engineering Capabilities](#engineering-capabilities) · [IDE Support](#ide-support) · [Deep Dive](#deep-dive)
 
 ## Why
 
@@ -48,7 +48,7 @@ asd ui        # Start background service (MCP Server + Dashboard) — IDE and MC
 
 > **First time:** Manually enable the `autosnippet` service in your IDE's MCP settings.
 
-> **Tip:** Stronger models work better. We recommend Claude Opus 4 / Sonnet 4, GPT-5, or Gemini 3 Pro in Cursor / Copilot for more accurate patterns and fewer false positives.
+> **Tip:** Stronger models work better. We recommend Claude Opus 4.6 / Sonnet 4.6, GPT-5.4, or Gemini 3.1 Pro in Cursor / Copilot for more accurate patterns and fewer false positives.
 
 ### Cold Start: Build Your Knowledge Base
 
@@ -145,7 +145,7 @@ Creativity at capability boundaries. Three progressive modes — Reuse (0ms) →
 4. **Signal-Driven > Time-Driven** — Trigger on signal saturation, not scheduled scans
 5. **Defense in Depth** — Constitution → Gateway → Permission → SafetyPolicy → PathGuard → ConfidenceRouter
 
-> Organ implementation details, engineering metrics, and defense chain breakdown in [Technical Reference](docs/technical-reference.en.md)
+> Organ implementation details, engineering metrics, and defense chain breakdown in [Technical Book](https://docs.gaoxuefeng.com/visual-tour)
 
 ---
 
@@ -240,22 +240,22 @@ All configuration auto-generated by `asd setup`. Run `asd upgrade` after updates
 
 ---
 
-## Docs
-
-| Document | Content |
-|----------|---------|
-| **[Technical Book](https://docs.gaoxuefeng.com/part1/ch01-introduction)** | **Deep dive into every module — architecture decisions, data flows, and implementation details** |
-| [Technical Reference](docs/technical-reference.en.md) | Six subsystem implementation details, engineering metrics, defense chain |
-| [Dashboard](docs/dashboard.en.md) | Dashboard views and tech stack |
-| [CLI Reference](docs/cli-reference.en.md) | Full usage for all 20 commands |
-| [MCP Tools Reference](docs/mcp-tools.en.md) | 16 MCP tools with parameters and usage |
-| [Architecture](docs/architecture.en.md) | Overall architecture and module design |
-| [Configuration](docs/configuration.en.md) | LLM and advanced configuration options |
-| [Guard Guide](docs/guard.en.md) | Guard rules and customization |
-| [IDE Integration](docs/ide-integration.en.md) | Detailed IDE setup guides |
-| [Lark Integration](docs/lark-integration.en.md) | Lark Bot setup and usage |
-| [Agent Architecture](docs/agent-architecture.en.md) | Agent Runtime and MCP protocol |
-| [Development](docs/development.en.md) | Contributing and dev environment setup |
+## Deep Dive
+
+> **[Visual Tour — Understand the entire system in 5 minutes](https://docs.gaoxuefeng.com/visual-tour)** · 25 hand-drawn architecture diagrams from workflow to Agent loop
+
+| Chapter | Content |
+|---------|--------|
+| [Introduction](https://docs.gaoxuefeng.com/part1/ch01-introduction) | Problem definition, solution overview, quick start |
+| [SOUL Principles](https://docs.gaoxuefeng.com/part1/ch02-soul) | 3 hard constraints + 5 design philosophies |
+| [Architecture](https://docs.gaoxuefeng.com/part2/ch03-architecture) | 7-layer DDD with module topology |
+| [Security Pipeline](https://docs.gaoxuefeng.com/part2/ch04-security) | Six-layer defense in depth |
+| [Code Understanding](https://docs.gaoxuefeng.com/part2/ch05-ast) | 10-language Tree-sitter AST analysis |
+| [Knowledge Domain](https://docs.gaoxuefeng.com/part3/ch06-knowledge-entry) | Unified entity, lifecycle, quality scoring |
+| [Core Services](https://docs.gaoxuefeng.com/part4/ch09-bootstrap) | Bootstrap, Guard, Search, Metabolism |
+| [Agent Intelligence](https://docs.gaoxuefeng.com/part5/ch13-agent-runtime) | ReAct loop, orthogonal composition, 61+ tools |
+| [Platform & Delivery](https://docs.gaoxuefeng.com/part6/ch16-infrastructure) | Data infrastructure, MCP, four-interface access |
+| [BiliDili Cold Start](https://docs.gaoxuefeng.com/part7/ch19-bilidili-coldstart) | Real data: 8.4M tokens, 101 candidates |
 
 ---
 
@@ -265,6 +265,31 @@ All configuration auto-generated by `asd setup`. Run `asd upgrade` after updates
 - macOS recommended (Xcode features require it; other features are cross-platform)
 - better-sqlite3 (bundled)
 
+### Recommended: Local Embedding for Semantic Search
+
+AutoSnippet has a built-in hybrid search engine (keyword + vector semantic). Install a local embedding model to unlock semantic search — concept-level matching that finds relevant recipes even when exact keywords don't match.
+
+```bash
+# Install Ollama (https://ollama.com)
+brew install ollama && ollama serve
+
+# Pull the recommended model (~639MB, supports Chinese + English + code)
+ollama pull qwen3-embedding:0.6b
+```
+
+Then add to your project's `.env`:
+
+```bash
+ASD_EMBED_PROVIDER=ollama
+ASD_EMBED_MODEL=qwen3-embedding:0.6b
+```
+
+Or configure it in Dashboard (`asd ui`) → Settings → Embedding Model.
+
+After configuring, run `asd embed` to build the vector index. Semantic search adds ~200–400ms per query (local inference, no API calls, no data leaves your machine).
+
+> **Without a local model**, search still works — it uses field-weighted keyword matching, which is fast and accurate for exact terms. Semantic search is a bonus layer for concept-level queries like *"how to avoid data races"* or *"cookie persistence"*.
+
 ## Contributing
 
 1. Run `npm test` before submitting

package/config/default.json

@@ -39,7 +39,7 @@
   },
   "ai": {
     "provider": "openai",
-    "model": "gpt-4",
+    "model": "gpt-5.4",
     "temperature": 0.7,
     "maxTokens": 2000
   },