Add Code Review Assistant chat app (Hierarchical Delegation)

examples/chat-apps/README.md

@@ -8,6 +8,7 @@ Interactive multi-agent chat applications with Streamlit UI for testing AgentQ o
 |-----|---------|-------------|
 | [`support-bot/`](support-bot/) | **Router / Dispatcher** | Customer support bot that routes questions to specialist agents (Billing, Technical, FAQ). Demonstrates branching trace topology. |
 | [`research-assistant/`](research-assistant/) | **Sequential Pipeline** | Research assistant that flows queries through Researcher → Analyzer → Writer agents. Demonstrates linear trace chains. |
+| [`code-review-assistant/`](code-review-assistant/) | **Hierarchical Delegation** | Code review assistant where a Manager delegates to Security, Style, and Logic reviewer agents. Demonstrates parent-child trace hierarchy. |
 
 ## Prerequisites
 
@@ -56,7 +57,11 @@ examples/chat-apps/
 │   ├── README.md
 │   ├── requirements.txt
 │   └── main.py
-└── research-assistant/
+├── research-assistant/
+│   ├── README.md
+│   ├── requirements.txt
+│   └── main.py
+└── code-review-assistant/
     ├── README.md
     ├── requirements.txt
     └── main.py

examples/chat-apps/code-review-assistant/README.md

@@ -0,0 +1,72 @@
+# Code Review Assistant — Hierarchical Delegation Pattern
+
+A code review assistant where a Manager agent delegates to specialist reviewer agents. Demonstrates **hierarchical parent-child trace topology** — the Manager is the parent span, each reviewer is a child span with its own tool + LLM sub-spans.
+
+## Architecture
+
+```
+User Pastes Code
+       │
+       ▼
+┌──────────────┐
+│   Manager    │  ← Orchestrates the review, consolidates findings
+│   Agent      │
+└──┬───┬───┬───┘
+   │   │   │
+   ▼   ▼   ▼
+┌────┐┌────┐┌────┐
+│ 🔒 ││ 🎨 ││ 🧠 │
+│Sec.││Sty.││Log.│
+│Rev.││Rev.││Rev.│
+└────┘└────┘└────┘
+   │   │   │
+   └───┼───┘
+       ▼
+  Consolidated
+    Report
+```
+
+Each reviewer runs its own static analysis tools and LLM-based code inspection, all appearing as child spans under the Manager in AgentQ traces.
+
+## Run
+
+```bash
+# Create a virtual environment
+python -m venv .venv && source .venv/bin/activate
+
+# Install dependencies
+pip install -r requirements.txt
+
+# Start the app
+streamlit run main.py
+```
+
+Then open `http://localhost:3000` to see traces in the AgentQ dashboard.
+
+## What to Try
+
+- **Code with `eval()` or SQL queries** → Security reviewer flags injection/execution risks
+- **Code with hardcoded passwords** → Security reviewer flags credential exposure
+- **Code with classes or complex imports** → Style reviewer provides specific guidance
+- **Code with nested if/else** → Logic reviewer flags conditional complexity
+- **Code with loops** → Logic reviewer analyzes efficiency and termination
+- **Any Python code** → All three reviewers provide a general review
+
+Use the **"Show individual reviewer reports"** expander to see each specialist's detailed findings alongside their tool outputs.
+
+## Trace Topology
+
+```
+session (conversation)
+  └── manager-agent
+        ├── security-reviewer
+        │     ├── static-analysis-scan (tool)
+        │     └── analyze-security (LLM)
+        ├── style-reviewer
+        │     ├── lint-check (tool)
+        │     └── analyze-style (LLM)
+        ├── logic-reviewer
+        │     ├── complexity-analysis (tool)
+        │     └── analyze-logic (LLM)
+        └── synthesize-report (LLM)
+```