update project structure

Cargo.toml

@@ -1,36 +1,7 @@
-[package]
-name = "open-web-agent-rs"
-version = "0.1.0"
-edition = "2021"
-license = "MIT"
+[workspace]
+members = [
+    "crates/agent-server",
+    "crates/local_inference_engine",
+]
 
-[[bin]]
-edition = "2021"
-name = "agent-server"
-path = "src/main.rs"
-
-[dependencies]
-axum = { version = "0.8", features = ["multipart"] }
-serde = { version = "1.0", features = ["derive"] }
-tokio = { version = "1.0", features = ["full"] }
-tracing = "0.1"
-tracing-subscriber = { version = "0.3", features = ["env-filter"] }
-http = "1.1.0"
-tokio-stream = "0.1.16"
-uuid = { version = "1.11.0", features = ["v4"] }
-tokio-util = { version = "0.7", features = ["io"] }
-serde_json = "1.0.133"
-futures = "0.3.31"
-dotenv = "0.15.0"
-shell-escape = "0.1.5"
-rust-embed = "8.5.0"
-bytes = "1.8.0"
-lazy_static = "1.5.0"
-sled = "0.34.7"
-tower-http = { version = "0.6.2", features = ["trace", "cors"] }
-tower = "0.5.2"
-anyhow = "1.0.97"
-base64 = "0.22.1"
-fips204 = "0.4.6"
-rmcp = { git = "https://github.com/modelcontextprotocol/rust-sdk", branch = "main", features = ["server", "transport-streamable-http-server",    "transport-sse-server", "transport-io",] }
-mime_guess = "2.0.5"
+resolver = "2"

Local.Dockerfile

@@ -14,11 +14,11 @@ RUN rustup target add aarch64-unknown-linux-musl
 
 # Copy only necessary files for building
 COPY Cargo.toml Cargo.lock ./
-COPY src ./src
+COPY crates ./crates
 COPY assets ./assets
 
 # Build with musl target for static linking
-RUN cargo build --release --target aarch64-unknown-linux-musl && \
+RUN cargo build -p agent-server --release --target aarch64-unknown-linux-musl && \
     strip /build-context/target/aarch64-unknown-linux-musl/release/agent-server
 
 # Stage 2: Build Bun dependencies

README.md

@@ -1,12 +1,69 @@
 # open-web-agent-rs
 
-A Rust-based web agent with an embedded openai compatible inference server (supports gemma models only).
+A Rust-based web agent with an embedded OpenAI-compatible inference server (supports Gemma models only).
 
-## Quickstart
+## Project Structure
+
+This project is organized as a Cargo workspace with the following crates:
+
+- `agent-server`: The main web agent server
+- `local_inference_engine`: An embedded OpenAI-compatible inference server for Gemma models
+
+## Setup
+
+1. Clone the repository
+2. Copy the example environment file:
    ```bash
    cp .env.example .env
+   ```
+3. Install JavaScript dependencies:
+   ```bash
    bun i
-(cd local_inference_server && cargo run --release -- --server)
+   ```
+4. Start the SearXNG search engine:
+   ```bash
    docker compose up -d searxng
+   ```
+
+## Running the Project
+
+### Local Inference Engine
+
+To run the local inference engine:
+
+```bash
+cd crates/local_inference_engine
+cargo run --release -- --server
+```
+
+### Agent Server
+
+To run the agent server:
+
+```bash
+cargo run -p agent-server
+```
+
+### Development Mode
+
+For development with automatic reloading:
+
+```bash
 bun dev
 ```
+
+## Building
+
+To build all crates in the workspace:
+
+```bash
+cargo build
+```
+
+To build a specific crate:
+
+```bash
+cargo build -p agent-server
+# or
+cargo build -p local_inference_engine
+```

Remote.Dockerfile

@@ -14,11 +14,11 @@ RUN rustup target add x86_64-unknown-linux-musl
 
 # Copy only necessary files for building
 COPY Cargo.toml Cargo.lock ./
-COPY src ./src
+COPY crates ./crates
 COPY assets ./assets
 
 # Build with musl target for static linking
-RUN cargo build --release --target x86_64-unknown-linux-musl && \
+RUN cargo build -p agent-server --release --target x86_64-unknown-linux-musl && \
     strip /build-context/target/x86_64-unknown-linux-musl/release/agent-server
 
 # Stage 2: Build Bun dependencies

crates/agent-server/Cargo.toml

@@ -0,0 +1,36 @@
+[package]
+name = "agent-server"
+version = "0.1.0"
+edition = "2021"
+license = "MIT"
+
+[[bin]]
+edition = "2021"
+name = "agent-server"
+path = "src/main.rs"
+
+[dependencies]
+axum = { version = "0.8", features = ["multipart"] }
+serde = { version = "1.0", features = ["derive"] }
+tokio = { version = "1.0", features = ["full"] }
+tracing = "0.1"
+tracing-subscriber = { version = "0.3", features = ["env-filter"] }
+http = "1.1.0"
+tokio-stream = "0.1.16"
+uuid = { version = "1.11.0", features = ["v4"] }
+tokio-util = { version = "0.7", features = ["io"] }
+serde_json = "1.0.133"
+futures = "0.3.31"
+dotenv = "0.15.0"
+shell-escape = "0.1.5"
+rust-embed = "8.5.0"
+bytes = "1.8.0"
+lazy_static = "1.5.0"
+sled = "0.34.7"
+tower-http = { version = "0.6.2", features = ["trace", "cors"] }
+tower = "0.5.2"
+anyhow = "1.0.97"
+base64 = "0.22.1"
+fips204 = "0.4.6"
+rmcp = { git = "https://github.com/modelcontextprotocol/rust-sdk", branch = "main", features = ["server", "transport-streamable-http-server",    "transport-sse-server", "transport-io",] }
+mime_guess = "2.0.5"

crates/agent-server/src/routes.rs

@@ -13,7 +13,7 @@ use crate::agents::Agents;
 
 
 #[derive(Embed)]
-#[folder = "./node_modules/@modelcontextprotocol/inspector-client/dist"]
+#[folder = "../../node_modules/@modelcontextprotocol/inspector-client/dist"]
 struct Asset;
 
 pub struct StaticFile<T>(pub T);

crates/local_inference_engine/README.md

@@ -30,8 +30,7 @@ A Rust-based inference engine for running large language models locally. This to
 
 2. Build the local inference engine:
    ```bash
-   cd local_inference_engine
-   cargo build --release
+   cargo build -p local_inference_engine --release
    ```
 
 ## Usage
@@ -41,7 +40,7 @@ A Rust-based inference engine for running large language models locally. This to
 Run the inference engine in CLI mode to generate text directly:
 
 ```bash
-cargo run --release -- --prompt "Your prompt text here" --which 3-1b-it
+cargo run -p local_inference_engine --release -- --prompt "Your prompt text here" --which 3-1b-it
 ```
 
 #### CLI Options
@@ -63,7 +62,7 @@ cargo run --release -- --prompt "Your prompt text here" --which 3-1b-it
 Run the inference engine in server mode to expose an OpenAI-compatible API:
 
 ```bash
-cargo run --release -- --server --port 3777 --which 3-1b-it
+cargo run -p local_inference_engine --release -- --server --port 3777 --which 3-1b-it
 ```
 
 This starts a web server on the specified port (default: 3777) with an OpenAI-compatible chat completions endpoint.

package.json

@@ -7,7 +7,7 @@
   "private": true,
   "scripts": {
     "clean": "rm -rf .genaiscript && rm -rf dist && rm -rf node_modules && rm -rf open-web-agent-rs && rm -rf target && rm -rf packages/genaiscript-rust-shim/dist",
-    "dev": "bun i && ./killport.js 3006 && bun run build && cargo watch -x 'run src/main.rs'",
+    "dev": "bun i && ./killport.js 3006 && bun run build && cargo watch -x 'run -p agent-server'",
     "start": "docker compose up --build",
     "ai:search": "genaiscript run packages/genaiscript/genaisrc/web-search.genai.mts --vars USER_INPUT='who won the 2024 election?'",
     "shim:ai:search": "pnpm build && ./packages/genaiscript-rust-shim/dist/genaiscript-rust-shim.js --file=packages/genaiscript/genaisrc/web-search.genai.mts USER_INPUT=\"Who won the 2024 presidential election?\"\n",
@@ -17,7 +17,7 @@
     "prod:logs": "fly logs",
     "test-http": "test/test-search.ts",
     "mcp-inspector": "bunx @modelcontextprotocol/inspector",
-    "build": "(cd packages/genaiscript-rust-shim && bun run buildShim && bun run setupDev && cargo build)"
+    "build": "(cd packages/genaiscript-rust-shim && bun run buildShim && bun run setupDev && cargo build -p agent-server)"
   },
   "dependencies": {
     "@modelcontextprotocol/inspector": "^0.14.0"