Add MiniMax image generation tool

.env.example

@@ -30,6 +30,9 @@ MODEL_ID=claude-sonnet-4-6
 # MODEL_ID=MiniMax-M3
 # MODEL_ID=MiniMax-M2.7
 # MODEL_ID=MiniMax-M2.7-highspeed
+# Image generation tool (optional; s02)
+# MINIMAX_API_KEY=sk-xxx
+# MINIMAX_API_HOST=https://api.minimax.io
 
 # GLM (Zhipu)      https://z.ai
 # ANTHROPIC_BASE_URL=https://api.z.ai/api/anthropic
@@ -53,6 +56,9 @@ MODEL_ID=claude-sonnet-4-6
 # MODEL_ID=MiniMax-M3
 # MODEL_ID=MiniMax-M2.7
 # MODEL_ID=MiniMax-M2.7-highspeed
+# Image generation tool (optional; s02)
+# MINIMAX_API_KEY=sk-xxx
+# MINIMAX_API_HOST=https://api.minimaxi.com
 
 # GLM (Zhipu)      https://open.bigmodel.cn
 # ANTHROPIC_BASE_URL=https://open.bigmodel.cn/api/anthropic

s02_tool_use/README.en.md

@@ -30,7 +30,7 @@ Adding a tool to the Agent requires just two things:
 
 ---
 
-## From 1 Tool to 5 Tools
+## From 1 Tool to 6 Tools
 
 s01 had only bash:
 
@@ -40,7 +40,7 @@ TOOLS = [{"name": "bash", ...}]
 def run_bash(command): ...
 ```
 
-s02 expands to 5 tools, each independently defined:
+s02 expands to 6 tools, each independently defined:
 
 ```python
 TOOLS = [
@@ -49,6 +49,7 @@ TOOLS = [
     {"name": "write_file", "description": "Write content to file.", ...},
     {"name": "edit_file",  "description": "Replace text in file once.", ...},
     {"name": "glob",       "description": "Find files by pattern.", ...},
+    {"name": "generate_image", "description": "Generate an image from a text prompt.", ...},
 ]
 ```
 
@@ -75,6 +76,9 @@ def run_edit(path, old_text, new_text):
 def run_glob(pattern):
     import glob as g
     return "\n".join(g.glob(pattern, root_dir=WORKDIR))
+
+def run_generate_image(prompt, aspect_ratio="1:1"):
+    ...
 ```
 
 ---
@@ -88,6 +92,7 @@ TOOL_HANDLERS = {
     "write_file": run_write,
     "edit_file":  run_edit,
     "glob":       run_glob,
+    "generate_image": run_generate_image,
 }
 
 # Only one line changed in the loop — from hardcoded run_bash to dispatch lookup:
@@ -125,7 +130,7 @@ The teaching version executes them one by one in the original `response.content`
 
 | Component | Before (s01) | After (s02) |
 |-----------|-------------|-------------|
-| Tool count | 1 (bash) | 5 (+read, write, edit, glob) |
+| Tool count | 1 (bash) | 6 (+read, write, edit, glob, generate_image) |
 | Tool execution | Hardcoded `run_bash()` | TOOL_HANDLERS dispatch lookup |
 | Path safety | None | safe_path validation (file tools only) |
 | Loop | `while True` + `stop_reason` | Identical to s01 |
@@ -139,20 +144,28 @@ cd learn-claude-code
 python s02_tool_use/code.py
 ```
 
+The optional `generate_image` tool also needs these values in `.env`. Use `https://api.minimaxi.com` as the host for China mainland.
+
+```sh
+MINIMAX_API_KEY=sk-xxx
+MINIMAX_API_HOST=https://api.minimax.io
+```
+
 Try these prompts:
 
 1. `Read the file README.md and tell me what this project is about`
 2. `Create a file called test.py that prints "hello", then read it back`
 3. `Find all Python files in this directory`
 4. `Read both README.md and requirements.txt, then create a summary file`
+5. `Generate a 16:9 image of a coding workspace at sunrise`
 
 What to watch for: When does the model call just one tool, and when does it call multiple at once? Are multiple tool calls executed in the correct order?
 
 ---
 
 ## What's Next
 
-The Agent now has 5 specialized tools. File tools are protected by `safe_path`, but bash is unrestricted — `rm -rf /` still runs.
+The Agent now has 6 specialized tools. File tools are protected by `safe_path`, but bash is unrestricted — `rm -rf /` still runs.
 
 → s03 Permission: Add a gate before tool execution — is this operation safe? Does it need user approval?
 

s02_tool_use/README.ja.md

@@ -30,7 +30,7 @@ Agent にツールを追加するには、たった二つ：
 
 ---
 
-## 1 つのツールから 5 つのツールへ
+## From 1 Tool to 6 Tools
 
 s01 には bash だけだった：
 
@@ -40,7 +40,7 @@ TOOLS = [{"name": "bash", ...}]
 def run_bash(command): ...
 ```
 
-s02 では 5 つに増え、各ツールは独立して定義される：
+s02 expands to 6 tools, each independently defined:
 
 ```python
 TOOLS = [
@@ -49,6 +49,7 @@ TOOLS = [
     {"name": "write_file", "description": "Write content to file.", ...},
     {"name": "edit_file",  "description": "Replace text in file once.", ...},
     {"name": "glob",       "description": "Find files by pattern.", ...},
+    {"name": "generate_image", "description": "Generate an image from a text prompt.", ...},
 ]
 ```
 
@@ -75,6 +76,9 @@ def run_edit(path, old_text, new_text):
 def run_glob(pattern):
     import glob as g
     return "\n".join(g.glob(pattern, root_dir=WORKDIR))
+
+def run_generate_image(prompt, aspect_ratio="1:1"):
+    ...
 ```
 
 ---
@@ -88,6 +92,7 @@ TOOL_HANDLERS = {
     "write_file": run_write,
     "edit_file":  run_edit,
     "glob":       run_glob,
+    "generate_image": run_generate_image,
 }
 
 # ループ内で変更されたのは一行だけ — ハードコードの run_bash から検索ディスパッチへ：
@@ -125,7 +130,7 @@ for block in response.content:
 
 | コンポーネント | 変更前 (s01) | 変更後 (s02) |
 |--------------|-------------|-------------|
-| ツール数 | 1 (bash) | 5 (+read, write, edit, glob) |
+| Tool count | 1 (bash) | 6 (+read, write, edit, glob, generate_image) |
 | ツール実行 | ハードコード `run_bash()` | TOOL_HANDLERS 検索ディスパッチ |
 | パス安全性 | なし | safe_path 検証（file tools のみ） |
 | ループ | `while True` + `stop_reason` | s01 と完全に同一 |
@@ -139,20 +144,28 @@ cd learn-claude-code
 python s02_tool_use/code.py
 ```
 
+The optional `generate_image` tool also needs these values in `.env`. Use `https://api.minimaxi.com` as the host for China mainland.
+
+```sh
+MINIMAX_API_KEY=sk-xxx
+MINIMAX_API_HOST=https://api.minimax.io
+```
+
 以下のプロンプトを試してみよう：
 
 1. `Read the file README.md and tell me what this project is about`
 2. `Create a file called test.py that prints "hello", then read it back`
 3. `Find all Python files in this directory`
 4. `Read both README.md and requirements.txt, then create a summary file`
+5. `Generate a 16:9 image of a coding workspace at sunrise`
 
 観察のポイント：モデルがツールを一つだけ呼び出すときと、複数同時に呼び出すときの違い。複数のツール呼び出しは正しい順序で実行されているか？
 
 ---
 
 ## 次へ
 
-Agent は 5 つの専用ツールを持つようになった。file tools は `safe_path` で保護されるが、bash は制限なし — `rm -rf /` はまだ実行できる。
+The Agent now has 6 specialized tools. File tools are protected by `safe_path`, but bash remains unrestricted: `rm -rf /` still runs.
 
 → s03 Permission：ツール実行前にゲートを追加 — この操作は安全か？ ユーザーの承認が必要か？
 

s02_tool_use/README.md

@@ -30,7 +30,7 @@ s01 的循环完全保留（LLM 调用、stop_reason 判断、消息追加）。
 
 ---
 
-## 从 1 个工具到 5 个工具
+## From 1 Tool to 6 Tools
 
 s01 只有一个 bash：
 
@@ -40,7 +40,7 @@ TOOLS = [{"name": "bash", ...}]
 def run_bash(command): ...
 ```
 
-s02 加到 5 个，每个工具都是独立定义：
+s02 expands to 6 tools, each independently defined:
 
 ```python
 TOOLS = [
@@ -49,6 +49,7 @@ TOOLS = [
     {"name": "write_file", "description": "Write content to file.", ...},
     {"name": "edit_file",  "description": "Replace text in file once.", ...},
     {"name": "glob",       "description": "Find files by pattern.", ...},
+    {"name": "generate_image", "description": "Generate an image from a text prompt.", ...},
 ]
 ```
 
@@ -75,6 +76,9 @@ def run_edit(path, old_text, new_text):
 def run_glob(pattern):
     import glob as g
     return "\n".join(g.glob(pattern, root_dir=WORKDIR))
+
+def run_generate_image(prompt, aspect_ratio="1:1"):
+    ...
 ```
 
 ---
@@ -88,6 +92,7 @@ TOOL_HANDLERS = {
     "write_file": run_write,
     "edit_file":  run_edit,
     "glob":       run_glob,
+    "generate_image": run_generate_image,
 }
 
 # 循环里只改了一行——从硬编码 run_bash 变成查表：
@@ -125,7 +130,7 @@ for block in response.content:
 
 | 组件 | 之前 (s01) | 之后 (s02) |
 |------|-----------|-----------|
-| 工具数量 | 1 (bash) | 5 (+read, write, edit, glob) |
+| Tool count | 1 (bash) | 6 (+read, write, edit, glob, generate_image) |
 | 工具执行 | 硬编码 `run_bash()` | TOOL_HANDLERS 查表分发 |
 | 路径安全 | 无 | safe_path 校验（仅 file tools） |
 | 循环 | `while True` + `stop_reason` | 与 s01 完全一致 |
@@ -139,20 +144,28 @@ cd learn-claude-code
 python s02_tool_use/code.py
 ```
 
+The optional `generate_image` tool also needs these values in `.env`. Use `https://api.minimaxi.com` as the host for China mainland.
+
+```sh
+MINIMAX_API_KEY=sk-xxx
+MINIMAX_API_HOST=https://api.minimax.io
+```
+
 试试这些 prompt：
 
 1. `Read the file README.md and tell me what this project is about`
 2. `Create a file called test.py that prints "hello", then read it back`
 3. `Find all Python files in this directory`
 4. `Read both README.md and requirements.txt, then create a summary file`
+5. `Generate a 16:9 image of a coding workspace at sunrise`
 
 观察重点：模型什么时候只调一个工具，什么时候一次调多个？多个工具调用的顺序和结果是否正确？
 
 ---
 
 ## 接下来
 
-现在 Agent 有 5 个专用工具。file tools 受 `safe_path` 保护，但 bash 不受限制，`rm -rf /` 还是能跑。
+The Agent now has 6 specialized tools. File tools are protected by `safe_path`, but bash remains unrestricted: `rm -rf /` still runs.
 
 s03 Permission → 在工具执行之前加一道门：这个操作安全吗？需要用户批准吗？
 

s02_tool_use/code.py

@@ -1,19 +1,19 @@
 #!/usr/bin/env python3
 """
-s02: Tool Use — 在 s01 基础上新增 4 个工具 + 分发映射。
+s02: Tool Use - adds five tools and a dispatch map on top of s01.
 
 运行: python s02_tool_use/code.py
 需要: pip install anthropic python-dotenv + .env 中配置 ANTHROPIC_API_KEY
 
 本文件 = s01 的全部代码 + 以下新增:
-  + run_read / run_write / run_edit / run_glob 四个工具实现
+  + Five tool implementations: run_read / run_write / run_edit / run_glob / run_generate_image
   + TOOL_HANDLERS 分发映射（替代 s01 中硬编码的 run_bash 调用）
   + safe_path 路径安全校验
 
 循环本身（agent_loop）与 s01 完全一致。
 """
 
-import os, subprocess
+import os, subprocess, json, urllib.request
 from pathlib import Path
 
 try:
@@ -60,7 +60,7 @@ def run_bash(command: str) -> str:
 
 
 # ═══════════════════════════════════════════════════════════
-#  NEW in s02: 4 个新工具
+#  NEW in s02: 5 tools
 # ═══════════════════════════════════════════════════════════
 
 def safe_path(p: str) -> Path:
@@ -114,8 +114,44 @@ def run_glob(pattern: str) -> str:
         return f"Error: {e}"
 
 
+def run_generate_image(prompt: str, aspect_ratio: str = "1:1") -> str:
+    """Generate an image with MiniMax and return the image URL."""
+    api_key = os.getenv("MINIMAX_API_KEY")
+    if not api_key:
+        return "Error: set MINIMAX_API_KEY to use image generation"
+    api_host = (os.getenv("MINIMAX_API_HOST") or "https://api.minimax.io").rstrip("/")
+
+    payload = {
+        "model": "image-01",
+        "prompt": prompt,
+        "aspect_ratio": aspect_ratio,
+        "n": 1,
+        "response_format": "url",
+    }
+    req = urllib.request.Request(
+        f"{api_host}/v1/image_generation",
+        data=json.dumps(payload).encode("utf-8"),
+        headers={
+            "Authorization": f"Bearer {api_key}",
+            "Content-Type": "application/json",
+        },
+        method="POST",
+    )
+    try:
+        with urllib.request.urlopen(req, timeout=120) as resp:
+            data = json.loads(resp.read().decode("utf-8"))
+        status = data.get("base_resp", {}).get("status_code")
+        if status not in (None, 0):
+            msg = data.get("base_resp", {}).get("status_msg", "unknown error")
+            return f"Error: MiniMax image generation failed ({status}): {msg}"
+        urls = data.get("data", {}).get("image_urls") or []
+        return urls[0] if urls else f"Error: no image URL returned: {data}"
+    except Exception as e:
+        return f"Error: {e}"
+
+
 # ═══════════════════════════════════════════════════════════
-#  NEW in s02: 工具定义（s01 只有一个 bash，现在扩展到 5 个）
+#  NEW in s02: tool definitions (s01 has only bash; s02 has 6 tools)
 # ═══════════════════════════════════════════════════════════
 
 TOOLS = [
@@ -129,6 +165,8 @@ def run_glob(pattern: str) -> str:
      "input_schema": {"type": "object", "properties": {"path": {"type": "string"}, "old_text": {"type": "string"}, "new_text": {"type": "string"}}, "required": ["path", "old_text", "new_text"]}},
     {"name": "glob", "description": "Find files matching a glob pattern.",
      "input_schema": {"type": "object", "properties": {"pattern": {"type": "string"}}, "required": ["pattern"]}},
+    {"name": "generate_image", "description": "Generate an image from a text prompt using MiniMax.",
+     "input_schema": {"type": "object", "properties": {"prompt": {"type": "string", "maxLength": 1500}, "aspect_ratio": {"type": "string", "default": "1:1", "enum": ["1:1", "16:9", "4:3", "3:2", "2:3", "3:4", "9:16", "21:9"]}}, "required": ["prompt"]}},
 ]
 
 # ═══════════════════════════════════════════════════════════
@@ -137,7 +175,7 @@ def run_glob(pattern: str) -> str:
 
 TOOL_HANDLERS = {
     "bash": run_bash, "read_file": run_read, "write_file": run_write,
-    "edit_file": run_edit, "glob": run_glob,
+    "edit_file": run_edit, "glob": run_glob, "generate_image": run_generate_image,
 }
 
 
@@ -171,7 +209,7 @@ def agent_loop(messages: list):
 
 
 if __name__ == "__main__":
-    print("s02: Tool Use — 在 s01 基础上加了 4 个工具")
+    print("s02: Tool Use - adds 5 tools on top of s01")
     print("输入问题，回车发送。输入 q 退出。\n")
 
     history = []