prompt-copilot-cli 0.1.2__tar.gz → 0.1.5__tar.gz

{prompt_copilot_cli-0.1.2/prompt_copilot_cli.egg-info → prompt_copilot_cli-0.1.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prompt-copilot-cli
-Version: 0.1.2
+Version: 0.1.5
 Summary: A lightweight terminal-based coding agent with file tools, command execution, multimodal image support, and MCP integration.
 Author: Jason Li
 License-Expression: MIT

{prompt_copilot_cli-0.1.2 → prompt_copilot_cli-0.1.5}/main.py

@@ -220,6 +220,7 @@ def t(key: str, **kwargs: Any) -> str:
 
 
 UI_SYSTEM_LANGUAGE = "en"
+APPLICATION_VERSION = "0.1.5"
 DEFAULT_SYSTEM_PROMPT = t("system_prompt")
 WORKSPACE_DIR = ROOT / "workspace"
 LOG_DIR = ROOT / "logs"
@@ -227,7 +228,7 @@ LOG_FILE = LOG_DIR / "agent_runtime.log"
 DEFAULT_MAX_CHAT_COUNT = 5
 CHAT_MESSAGE_MAX_COUNT = 6
 CONFIG_SAVE_FILE_PATH = ROOT / "config.json"
-RE_ACTION_DELAY = 5 # unit: seconds
+RE_ACTION_DELAY = 6 # unit: seconds
 TOOL_SUBPROCESS_TIMEOUT = 60 * 60  # 1 hour in seconds
 DEFAULT_MODEL_CONFIG: dict[str, Any] = {
     "model": "",
@@ -550,14 +551,58 @@ def safe_parse_tool_args(raw_arguments: Any) -> dict[str, Any]:
     return {}
 
 
-def build_multimodal_user_message(text: str, image_path: str | os.PathLike[str]) -> dict[str, Any]:
+def build_multimodal_user_message(
+    text: str,
+    image_path: str | os.PathLike[str],
+    max_bytes: int | None = None,
+) -> dict[str, Any]:
     path = Path(image_path).expanduser()
     if not path.exists():
         raise FileNotFoundError(f"Image file not found: {path}")
 
     image_bytes = path.read_bytes()
-    encoded = base64.b64encode(image_bytes).decode("ascii")
     mime_type = "image/png" if path.suffix.lower() == ".png" else "image/jpeg"
+
+    if max_bytes is None:
+        max_bytes = 120_000
+
+    if max_bytes and len(image_bytes) > max_bytes:
+        try:
+            import io
+            from PIL import Image
+
+            with Image.open(path) as img:
+                img = img.convert("RGB")
+                candidate_bytes = image_bytes
+                candidate_mime = mime_type
+                width, height = img.size
+                for scale in [1.0, 0.8, 0.6, 0.4, 0.2, 0.1, 0.05, 0.03, 0.01]:
+                    resized = img.resize((max(16, int(width * scale)), max(16, int(height * scale))), Image.Resampling.LANCZOS)
+                    for quality in [85, 70, 55, 40, 25, 15, 10]:
+                        buffer = io.BytesIO()
+                        resized.save(buffer, format="JPEG", quality=quality, optimize=True)
+                        candidate_bytes = buffer.getvalue()
+                        candidate_mime = "image/jpeg"
+                        encoded_candidate = base64.b64encode(candidate_bytes).decode("ascii")
+                        if len(encoded_candidate) <= max_bytes:
+                            image_bytes = candidate_bytes
+                            mime_type = candidate_mime
+                            break
+                    if len(base64.b64encode(candidate_bytes).decode("ascii")) <= max_bytes:
+                        break
+                else:
+                    tiny = img.resize((32, 32), Image.Resampling.LANCZOS)
+                    buffer = io.BytesIO()
+                    tiny.save(buffer, format="JPEG", quality=15, optimize=True)
+                    image_bytes = buffer.getvalue()
+                    mime_type = "image/jpeg"
+                    if len(base64.b64encode(image_bytes).decode("ascii")) > max_bytes:
+                        image_bytes = b"\x00"
+                        mime_type = "image/jpeg"
+        except Exception:
+            image_bytes = image_bytes[:max(1, max_bytes // 2)]
+
+    encoded = base64.b64encode(image_bytes).decode("ascii")
     return {
         "role": "user",
         "content": [
@@ -572,8 +617,8 @@ def build_multimodal_user_message(text: str, image_path: str | os.PathLike[str])
     }
 
 
-def build_multimodal_prompt_from_image(user_text: str, image_path: str | os.PathLike[str]) -> list[dict[str, Any]]:
-    return [build_multimodal_user_message(user_text, image_path)]
+def build_multimodal_prompt_from_image(user_text: str, image_path: str | os.PathLike[str], max_bytes: int | None = None) -> list[dict[str, Any]]:
+    return [build_multimodal_user_message(user_text, image_path, max_bytes=max_bytes)]
 
 
 def resolve_execution_cwd(cwd: Any, fallback: str | os.PathLike[str] | None = None) -> str:
@@ -1066,9 +1111,7 @@ def execute_tool_call(tool_call: Any) -> dict[str, Any]:
             logger.info("read_image_as_base64 result: %s", result)
             return result
 
-        image_bytes = path.read_bytes()
-        encoded = base64.b64encode(image_bytes).decode("ascii")
-        result = {"status": "ok", "content": json.dumps({"path": str(path), "base64": encoded}, ensure_ascii=False)}
+        result = {"status": "ok", "content": json.dumps({"path": str(path), "message": "Image loaded into context for multimodal analysis."}, ensure_ascii=False)}
         logger.info("read_image_as_base64 result: %s", result)
         return result
 
@@ -1462,6 +1505,20 @@ def run_agent(client: OpenAI, model: str, system_prompt: str, session_store: Ses
                 show_stage(t("tool_call_finished"), end_calling_tips[:80])
                 show_tool_result(tc, result)
                 messages.append({"role": "tool", "tool_call_id": tc.id, "content": json.dumps(result, ensure_ascii=False)})
+
+                if tc.function.name == "read_image_as_base64":
+                    try:
+                        payload = json.loads(result.get("content", "{}")) if isinstance(result.get("content"), str) else result.get("content", {})
+                        image_path = payload.get("path") if isinstance(payload, dict) else None
+                        if image_path:
+                            image_message = build_multimodal_user_message(
+                                "The user attached an image for multimodal analysis. Please inspect it carefully.",
+                                image_path,
+                                max_bytes=120_000,
+                            )
+                            messages.append(image_message)
+                    except Exception:
+                        logger.exception("Failed to append multimodal image message")
                 # if recorder:
                 #     recorder.record_tool_result(tc.function.name, result)
             except KeyboardInterrupt:
@@ -1594,6 +1651,7 @@ def interactive_loop(client: OpenAI, model: str, system_prompt: str, session_sto
 
 def build_cli_parser() -> argparse.ArgumentParser:
     parser = argparse.ArgumentParser(description=t("cli_parser_description"))
+    parser.add_argument("-v", "--version", action="version", version=f"%(prog)s {APPLICATION_VERSION}")
     parser.add_argument("-t", "--task", help=t("task_argument_help"))
     parser.add_argument("-d", "--workdir", default=WORKSPACE_DIR, help=t("workdir_argument_help"))
     parser.add_argument("-l", "--lang", default="en", help="language code for localization (default: en)")

{prompt_copilot_cli-0.1.2 → prompt_copilot_cli-0.1.5/prompt_copilot_cli.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: prompt-copilot-cli
-Version: 0.1.2
+Version: 0.1.5
 Summary: A lightweight terminal-based coding agent with file tools, command execution, multimodal image support, and MCP integration.
 Author: Jason Li
 License-Expression: MIT

{prompt_copilot_cli-0.1.2 → prompt_copilot_cli-0.1.5}/prompt_copilot_cli.egg-info/SOURCES.txt

@@ -11,6 +11,7 @@ prompt_copilot_cli.egg-info/dependency_links.txt
 prompt_copilot_cli.egg-info/entry_points.txt
 prompt_copilot_cli.egg-info/requires.txt
 prompt_copilot_cli.egg-info/top_level.txt
+tests/test_cli_args.py
 tests/test_command_timeout.py
 tests/test_file_tools.py
 tests/test_image_tool.py

{prompt_copilot_cli-0.1.2 → prompt_copilot_cli-0.1.5}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "prompt-copilot-cli"
-version = "0.1.2"
+version = "0.1.5"
 description = "A lightweight terminal-based coding agent with file tools, command execution, multimodal image support, and MCP integration."
 readme = "README.md"
 license = "MIT"

prompt_copilot_cli-0.1.5/tests/test_cli_args.py

@@ -0,0 +1,37 @@
+import contextlib
+import io
+import sys
+import unittest
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+import main
+
+
+class CLITests(unittest.TestCase):
+    def test_version_flag_prints_application_version_and_exits(self) -> None:
+        parser = main.build_cli_parser()
+        output = io.StringIO()
+
+        with contextlib.redirect_stdout(output):
+            with self.assertRaises(SystemExit) as exc:
+                parser.parse_args(["--version"])
+
+        self.assertEqual(exc.exception.code, 0)
+        self.assertIn(main.APPLICATION_VERSION, output.getvalue())
+
+    def test_help_flag_prints_usage_and_exits(self) -> None:
+        parser = main.build_cli_parser()
+        output = io.StringIO()
+
+        with contextlib.redirect_stdout(output):
+            with self.assertRaises(SystemExit) as exc:
+                parser.parse_args(["--help"])
+
+        self.assertEqual(exc.exception.code, 0)
+        self.assertIn("usage:", output.getvalue().lower())
+
+
+if __name__ == "__main__":
+    unittest.main()

{prompt_copilot_cli-0.1.2 → prompt_copilot_cli-0.1.5}/tests/test_image_tool.py

@@ -24,9 +24,8 @@ class ImageToolTests(unittest.TestCase):
         result = main.execute_tool_call(DummyToolCall("read_image_as_base64", {"path": str(image_path)}))
 
         self.assertEqual(result["status"], "ok")
-        payload = main.json.loads(result["content"])
-        self.assertTrue(Path(payload["path"]).exists())
-        self.assertTrue(payload["base64"])
+        self.assertIn("loaded into context", result["content"].lower())
+        self.assertNotIn("base64", result["content"].lower())
 
 
 if __name__ == "__main__":

prompt_copilot_cli-0.1.5/tests/test_multimodal_messages.py

@@ -0,0 +1,38 @@
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+import main
+
+
+class MultimodalMessageTests(unittest.TestCase):
+    def test_build_multimodal_user_message_with_image(self) -> None:
+        image_path = Path("tests") / "fixtures" / "sample.png"
+        message = main.build_multimodal_user_message("describe this", image_path)
+
+        self.assertEqual(message["role"], "user")
+        self.assertEqual(message["content"][0]["type"], "text")
+        self.assertEqual(message["content"][1]["type"], "image_url")
+        self.assertIn("data:image/", message["content"][1]["image_url"]["url"])
+
+    def test_build_multimodal_user_message_compresses_image(self) -> None:
+        from PIL import Image
+
+        with tempfile.TemporaryDirectory() as tmpdir:
+            image_path = Path(tmpdir) / "large.png"
+            large_image = Image.new("RGB", (4000, 4000), color=(255, 0, 0))
+            large_image.save(image_path)
+            original_bytes = image_path.read_bytes()
+            message = main.build_multimodal_user_message("describe this", image_path, max_bytes=5000)
+
+            payload = message["content"][1]["image_url"]["url"]
+            encoded = payload.split(",", 1)[1]
+            self.assertLess(len(encoded), 5000)
+            self.assertLess(len(encoded), len(original_bytes) * 2)
+
+
+if __name__ == "__main__":
+    unittest.main()

prompt_copilot_cli-0.1.2/tests/test_multimodal_messages.py

@@ -1,22 +0,0 @@
-import sys
-import unittest
-from pathlib import Path
-
-sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
-
-import main
-
-
-class MultimodalMessageTests(unittest.TestCase):
-    def test_build_multimodal_user_message_with_image(self) -> None:
-        image_path = Path("tests") / "fixtures" / "sample.png"
-        message = main.build_multimodal_user_message("describe this", image_path)
-
-        self.assertEqual(message["role"], "user")
-        self.assertEqual(message["content"][0]["type"], "text")
-        self.assertEqual(message["content"][1]["type"], "image_url")
-        self.assertIn("data:image/png;base64", message["content"][1]["image_url"]["url"])
-
-
-if __name__ == "__main__":
-    unittest.main()