autonomous-app 0.3.41__py3-none-any.whl → 0.3.42__py3-none-any.whl

autonomous/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.3.41"
+__version__ = "0.3.42"
 
 from dotenv import load_dotenv
 

autonomous/ai/imageagent.py

@@ -21,7 +21,7 @@ class ImageAgent(BaseAgent):
         self,
         prompt,
         negative_prompt="",
-        aspect_ratio="2KPortrait",
+        aspect_ratio="Portrait",
         files=None,
     ):
         return self.get_client(

autonomous/ai/models/gemini.py

@@ -18,12 +18,14 @@ from autonomous.model.automodel import AutoModel
 
 class GeminiAIModel(AutoModel):
     _client = None
-    _text_model = "gemini-3-pro-preview"
-    _summary_model = "gemini-2.5-flash"
-    _image_model = "gemini-3-pro-image-preview"
-    _json_model = "gemini-3-pro-preview"
-    _stt_model = "gemini-3-pro-preview"
-    _tts_model = "gemini-2.5-flash-preview-tts"
+
+    # Model definitions
+    _text_model = "gemini-1.5-pro"
+    _summary_model = "gemini-1.5-flash"
+    _json_model = "gemini-1.5-pro"
+    _stt_model = "gemini-1.5-pro"
+    _image_model = "imagen-3.0-generate-001"
+    _tts_model = "gemini-2.0-flash-exp"
 
     messages = ListAttr(StringAttr(default=[]))
     name = StringAttr(default="agent")
@@ -37,6 +39,30 @@ class GeminiAIModel(AutoModel):
 
     MAX_FILES = 14
     MAX_SUMMARY_TOKEN_LENGTH = 10000
+
+    SAFETY_SETTINGS = [
+        types.SafetySetting(
+            category="HARM_CATEGORY_HATE_SPEECH",
+            threshold="BLOCK_NONE",
+        ),
+        types.SafetySetting(
+            category="HARM_CATEGORY_SEXUALLY_EXPLICIT",
+            threshold="BLOCK_NONE",
+        ),
+        types.SafetySetting(
+            category="HARM_CATEGORY_DANGEROUS_CONTENT",
+            threshold="BLOCK_NONE",
+        ),
+        types.SafetySetting(
+            category="HARM_CATEGORY_HARASSMENT",
+            threshold="BLOCK_NONE",
+        ),
+        types.SafetySetting(
+            category="HARM_CATEGORY_CIVIC_INTEGRITY",
+            threshold="BLOCK_NONE",
+        ),
+    ]
+
     VOICES = {
         "Zephyr": ["female"],
         "Puck": ["male"],
@@ -73,52 +99,39 @@ class GeminiAIModel(AutoModel):
     @property
     def client(self):
         if not self._client:
-            # log("=== Initializing Gemini AI Client ===", _print=True)
             self._client = genai.Client(api_key=os.environ.get("GOOGLEAI_KEY"))
-            # log("=== Gemini AI Client Initialized ===", _print=True)
         return self._client
 
     def _add_function(self, user_function):
-        # This function is now a bit more advanced to conform to the Tool Use schema
         tool_schema = {
             "name": user_function.get("name"),
             "description": user_function.get("description"),
             "parameters": user_function.get("parameters"),
         }
-
-        # Validate that the schema has a name, description, and parameters
         if not all(
             [tool_schema["name"], tool_schema["description"], tool_schema["parameters"]]
         ):
             raise ValueError(
                 "Tool schema must have a 'name', 'description', and 'parameters' field."
             )
-
         return tool_schema
 
     def _create_wav_header(
         self, raw_audio_bytes, channels=1, rate=24000, sample_width=2
     ):
-        """Creates an in-memory WAV file from raw PCM audio bytes."""
         buffer = io.BytesIO()
         with wave.open(buffer, "wb") as wav_file:
-            # Set audio parameters
             wav_file.setnchannels(channels)
             wav_file.setsampwidth(sample_width)
-            wav_file.setframerate(rate)  # 16,000 Hz sample rate
-
-            # Write the raw audio data
+            wav_file.setframerate(rate)
             wav_file.writeframes(raw_audio_bytes)
-
         buffer.seek(0)
         return buffer
 
     def _add_context(self, context):
-        # Create in-memory file
         context_data = (
             json.dumps(context, indent=2) if isinstance(context, dict) else str(context)
         )
-
         f = io.BytesIO(context_data.encode("utf-8"))
         f.name = f"context-{self.pk}"
         return self._add_files([{"name": f.name, "file": f}])
@@ -128,31 +141,28 @@ class GeminiAIModel(AutoModel):
         for f in file_list[: self.MAX_FILES]:
             fn = f["name"]
             try:
-                result = self.client.files.delete(name=fn)
-            except Exception as e:
-                pass
-                # log(f"No existing file to delete for {fn}: {e}", _print=True)
-            else:
+                self.client.files.delete(name=fn)
+            except Exception:
                 pass
-                # log(f"Deleting old version of {fn}: {result}", _print=True)
 
-            # If the content is raw bytes, wrap it in BytesIO
             file_content = f["file"]
-            if isinstance(file_content, bytes):
-                fileobj = io.BytesIO(file_content)
-            else:
-                fileobj = file_content
+            fileobj = (
+                io.BytesIO(file_content)
+                if isinstance(file_content, bytes)
+                else file_content
+            )
+
             uploaded_file = self.client.files.upload(
                 file=fileobj,
                 config={"mime_type": mime_type, "display_name": fn},
             )
             uploaded_files.append(uploaded_file)
 
-            # This ensures the file is 'ACTIVE' before you use it in a prompt.
             while uploaded_file.state.name == "PROCESSING":
                 time.sleep(0.5)
                 uploaded_file = self.client.get_file(uploaded_file.name)
-        self.file_refs = [f.name for f in self.client.files.list()]  # Update file_refs
+
+        self.file_refs = [f.name for f in self.client.files.list()]
         self.save()
         return uploaded_files
 
@@ -160,8 +170,8 @@ class GeminiAIModel(AutoModel):
         self, message, function, additional_instructions="", uri="", context={}
     ):
         function_definition = self._add_function(function)
-
         contents = [message]
+
         if context:
             contents.extend(self._add_context(context))
             additional_instructions += (
@@ -169,12 +179,7 @@ class GeminiAIModel(AutoModel):
             )
 
         if uri:
-            contents.append(
-                Part.from_uri(
-                    file_uri=uri,
-                    mime_type="application/json",
-                ),
-            )
+            contents.append(Part.from_uri(file_uri=uri, mime_type="application/json"))
             additional_instructions += "\nUse the provided uri file for reference\n"
 
         response = self.client.models.generate_content(
@@ -183,24 +188,20 @@ class GeminiAIModel(AutoModel):
             config=types.GenerateContentConfig(
                 system_instruction=f"{self.instructions}.{additional_instructions}",
                 tools=[types.Tool(function_declarations=[function_definition])],
-                tool_config={
-                    "function_calling_config": {
-                        "mode": "ANY",  # Force a function call
-                    }
-                },
+                tool_config=types.ToolConfig(
+                    function_calling_config=types.FunctionCallingConfig(mode="ANY")
+                ),
             ),
         )
 
-        # The response is now a ToolCall, not a JSON string
         try:
-            # log(response.candidates[0].content.parts[0].function_call, _print=True)
+            if not response.candidates or not response.candidates[0].content.parts:
+                return {}
+
             tool_call = response.candidates[0].content.parts[0].function_call
             if tool_call and tool_call.name == function["name"]:
                 return tool_call.args
             else:
-                log(
-                    "==== Model did not return a tool call or returned the wrong one. ===="
-                )
                 log(f"Response: {response.text}", _print=True)
                 return {}
         except Exception as e:
@@ -216,12 +217,7 @@ class GeminiAIModel(AutoModel):
             )
 
         if uri:
-            contents.append(
-                Part.from_uri(
-                    file_uri=uri,
-                    mime_type="application/json",
-                ),
-            )
+            contents.append(Part.from_uri(file_uri=uri, mime_type="application/json"))
 
         response = self.client.models.generate_content(
             model=self._text_model,
@@ -230,20 +226,14 @@ class GeminiAIModel(AutoModel):
             ),
             contents=contents,
         )
-
-        # log(results, _print=True)
-        # log("=================== END REPORT ===================", _print=True)
         return response.text
 
     def summarize_text(self, text, primer=""):
         primer = primer or self.instructions
-
         updated_prompt_list = []
-        # Find all words in the prompt
         words = re.findall(r"\w+", text)
-        # Split the words into chunks
+
         for i in range(0, len(words), self.MAX_SUMMARY_TOKEN_LENGTH):
-            # Join a chunk of words and add to the list
             updated_prompt_list.append(
                 " ".join(words[i : i + self.MAX_SUMMARY_TOKEN_LENGTH])
             )
@@ -252,40 +242,27 @@ class GeminiAIModel(AutoModel):
         for p in updated_prompt_list:
             response = self.client.models.generate_content(
                 model=self._summary_model,
-                config=types.GenerateContentConfig(
-                    system_instruction=f"{primer}",
-                ),
-                contents=text,
+                config=types.GenerateContentConfig(system_instruction=f"{primer}"),
+                contents=p,
             )
             try:
                 summary = response.candidates[0].content.parts[0].text
+                full_summary += summary + "\n"
             except Exception as e:
-                log(f"{type(e)}:{e}\n\n Unable to generate content ====", _print=True)
+                log(f"Summary Error: {e}", _print=True)
                 break
-            else:
-                full_summary += summary + "\n"
-        return summary
+        return full_summary
 
     def generate_transcription(
-        self,
-        audio_file,
-        prompt="Transcribe this audio clip",
-        display_name="audio.mp3",
+        self, audio_file, prompt="Transcribe this audio clip", display_name="audio.mp3"
     ):
         myfile = self.client.files.upload(
             file=io.BytesIO(audio_file),
-            config={
-                "mime_type": "audio/mp3",
-                "display_name": display_name,
-            },
+            config={"mime_type": "audio/mp3", "display_name": display_name},
         )
-
         response = self.client.models.generate_content(
             model=self._stt_model,
-            contents=[
-                prompt,
-                myfile,
-            ],
+            contents=[prompt, myfile],
         )
         return response.text
 
@@ -309,44 +286,59 @@ class GeminiAIModel(AutoModel):
                     speech_config=types.SpeechConfig(
                         voice_config=types.VoiceConfig(
                             prebuilt_voice_config=types.PrebuiltVoiceConfig(
-                                voice_name=voice,
+                                voice_name=voice
                             )
                         )
                     ),
                 ),
             )
             blob = response.candidates[0].content.parts[0].inline_data
-
-            # Create a WAV file in memory from the raw audio bytes
             wav_buffer = self._create_wav_header(blob.data)
-
-            # 2. Load the WAV audio using pydub, which will now correctly read the header
             audio_segment = AudioSegment.from_file(wav_buffer, format="wav")
-
-            # 3. Create a new in-memory buffer for the MP3 output
             mp3_buffer = io.BytesIO()
-
-            # 4. Export the audio segment directly to the in-memory buffer
             audio_segment.export(mp3_buffer, format="mp3")
-
-            # 5. Return the bytes from the buffer, not the filename
             return mp3_buffer.getvalue()
-
         except Exception as e:
-            log(
-                f"==== Error: Unable to generate audio ====\n{type(e)}:{e}", _print=True
-            )
-            # You can return a default empty byte string or re-raise the exception
+            log(f"==== Audio Gen Error: {e} ====", _print=True)
             raise e
 
-    def generate_image(
-        self,
-        prompt,
-        negative_prompt="",
-        files=None,
-        aspect_ratio="3:4",
-        image_size="2K",
-    ):
+    def _get_image_config(self, aspect_ratio_input):
+        """
+        Parses custom aspect ratio keys (e.g., '2KPortrait') into valid
+        Google Gemini API parameters for ratio and size.
+        """
+        # Default fallback
+        ratio = "1:1"
+        size = "1K"
+
+        # Logic Mapping
+        # Keys match what your App sends in ttrpgbase.py
+        if aspect_ratio_input == "2KPortrait":
+            ratio = "9:16"
+            size = "2K"  # <--- THIS WAS MISSING BEFORE
+        elif aspect_ratio_input == "Portrait":
+            ratio = "9:16"
+            size = "1K"
+        elif aspect_ratio_input == "Landscape":
+            ratio = "16:9"
+            size = "1K"
+        elif aspect_ratio_input == "4K":
+            ratio = "16:9"
+            size = "4K"
+        elif aspect_ratio_input == "4KPortrait":
+            ratio = "9:16"
+            size = "4K"
+        elif aspect_ratio_input == "2K":
+            ratio = "1:1"
+            size = "2K"
+
+        # Pass-through for standard inputs
+        elif aspect_ratio_input in ["1:1", "3:4", "4:3", "9:16", "16:9"]:
+            ratio = aspect_ratio_input
+
+        return ratio, size
+
+    def generate_image(self, prompt, negative_prompt="", files=None, aspect_ratio="2K"):
         image = None
         contents = [prompt]
 
@@ -355,51 +347,36 @@ class GeminiAIModel(AutoModel):
             contents.extend(filerefs)
 
         try:
-            # log(self._image_model, contents, _print=True)
+            # 1. Resolve Aspect Ratio AND Size
+            valid_ratio, valid_size = self._get_image_config(aspect_ratio)
+
+            # 2. Call API with correct parameters
             response = self.client.models.generate_content(
                 model=self._image_model,
                 contents=contents,
                 config=types.GenerateContentConfig(
-                    safety_settings=[
-                        types.SafetySetting(
-                            category=types.HarmCategory.HARM_CATEGORY_HATE_SPEECH,
-                            threshold=types.HarmBlockThreshold.BLOCK_NONE,
-                        ),
-                        types.SafetySetting(
-                            category=types.HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
-                            threshold=types.HarmBlockThreshold.BLOCK_NONE,
-                        ),
-                        types.SafetySetting(
-                            category=types.HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
-                            threshold=types.HarmBlockThreshold.BLOCK_NONE,
-                        ),
-                        types.SafetySetting(
-                            category=types.HarmCategory.HARM_CATEGORY_HARASSMENT,
-                            threshold=types.HarmBlockThreshold.BLOCK_NONE,
-                        ),
-                        types.SafetySetting(
-                            category=types.HarmCategory.HARM_CATEGORY_CIVIC_INTEGRITY,
-                            threshold=types.HarmBlockThreshold.BLOCK_NONE,
-                        ),
-                    ],
+                    safety_settings=self.SAFETY_SETTINGS,
                     image_config=types.ImageConfig(
-                        aspect_ratio=aspect_ratio,
-                        image_size=image_size,
+                        aspect_ratio=valid_ratio,
+                        image_size=valid_size,  # Now passing "2K" or "4K" correctly
                     ),
                 ),
             )
-            log(response, _print=True)
-            log(response.candidates, _print=True)
-            image_parts = [
-                part.inline_data.data
-                for part in response.candidates[0].content.parts
-                if part.inline_data
-            ]
-            image = image_parts[0]
+
+            # 3. Extract Image Data
+            if response.candidates and response.candidates[0].content.parts:
+                for part in response.candidates[0].content.parts:
+                    if part.inline_data:
+                        image = part.inline_data.data
+                        break
+
+            if not image:
+                raise ValueError(
+                    f"API returned Success but no image data found. Response: {response}"
+                )
+
         except Exception as e:
-            log(
-                f"==== Error: Unable to create image ====\n\n{e}",
-                _print=True,
-            )
+            log(f"==== Error: Unable to create image ====\n\n{e}", _print=True)
             raise e
+
         return image

autonomous/ai/models/local_model.py

@@ -269,7 +269,9 @@ class LocalAIModel(AutoModel):
             "16:9": (1216, 832),
             "2K": (2048, 1080),
             "2KPortrait": (1080, 2048),
+            "Portrait": (1080, 2048),
             "4K": (3840, 2160),
+            "Landscape": (3840, 2160),
             "4KPortrait": (2160, 3840),
             "9:16": (832, 1216),
             "3:2": (1216, 832),

{autonomous_app-0.3.41.dist-info → autonomous_app-0.3.42.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: autonomous-app
-Version: 0.3.41
+Version: 0.3.42
 Summary: Containerized application framework built on Flask with additional libraries and tools for rapid development of web applications.
 Author-email: Steven A Moore <samoore@binghamton.edu>
 Project-URL: homepage, https://github.com/Sallenmoore/autonomous

{autonomous_app-0.3.41.dist-info → autonomous_app-0.3.42.dist-info}/RECORD

@@ -1,15 +1,15 @@
-autonomous/__init__.py,sha256=2Py6rnpUfu4t8c9Y9n3oQLD2-RGCbghqEkVltb7JTpw,95
+autonomous/__init__.py,sha256=XvTINoMxlBE_wUWlNeWGg7W7PDm0SpflzDGROfnBvEA,95
 autonomous/cli.py,sha256=z4AaGeWNW_uBLFAHng0J_lfS9v3fXemK1PeT85u4Eo4,42
 autonomous/logger.py,sha256=NQtgEaTWNAWfLSgqSP7ksXj1GpOuCgoUV711kSMm-WA,2022
 autonomous/ai/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autonomous/ai/audioagent.py,sha256=aZ25eEdze8S060-a4S0k319tgyl2aDTUa8dJu07mXn0,1092
 autonomous/ai/baseagent.py,sha256=icOPygr1NdH64u1ZYbwHHywYIY1ZtaLY9HtfNmUbx4k,4702
-autonomous/ai/imageagent.py,sha256=5eb4irUcWOD1HnquHX3jOTjOsed5Yn98rxSol84Lf18,902
+autonomous/ai/imageagent.py,sha256=1RT7OYTnRUo3q5k5w83A3cOh3hXUlrx0jRkg0YJSgZ0,900
 autonomous/ai/jsonagent.py,sha256=DNfZHMVCfc5nrkWJm2OebTYDkBwm_ZCeVWGIFGjB_Cg,1208
 autonomous/ai/textagent.py,sha256=0y2Hvb9pup1OnsA51hGPcD8yllZOZtztDLQvCNYABaw,1043
 autonomous/ai/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autonomous/ai/models/gemini.py,sha256=eu48gywNFpUFaqBt-4MFX2oRM5IED9rUTgtavM_HRG0,14468
-autonomous/ai/models/local_model.py,sha256=HjHDZqmdhlT29e5Q7GXNbZYbM9GVNiSzEJyBDpXWxtU,12431
+autonomous/ai/models/gemini.py,sha256=jCfOovhSBG0G9rg2DQ3plRV66VKPzW7w56UVjuMoXvU,13106
+autonomous/ai/models/local_model.py,sha256=uvevUIQgCwwrrFub-LoV4vbk71psdEPEzfzmuztn3sI,12508
 autonomous/apis/version_control/GHCallbacks.py,sha256=AyiUlYfV5JePi11GVyqYyXoj5UTbPKzS-HRRI94rjJo,1069
 autonomous/apis/version_control/GHOrganization.py,sha256=mi2livdsGurKiifbvuLwiFbdDzL77IlEfhwEa-tG77I,1155
 autonomous/apis/version_control/GHRepo.py,sha256=hTFHMkxSbSlVELfh8S6mq6ijkIKPRQO-Q5775ZjRKD4,4622
@@ -55,7 +55,7 @@ autonomous/taskrunner/__init__.py,sha256=ughX-QfWBas5W3aB2SiF887SWJ3Dzc2X43Yxtmp
 autonomous/taskrunner/autotasks.py,sha256=2zRaqHYqfdlgC_BQm6B6D2svN1ukyWeJJHwweZFHVoo,2616
 autonomous/taskrunner/task_router.py,sha256=W09HtRUuhwlnGxM5w4l6Hzw6mfS6L4ljWiMzD3ZVFeU,601
 autonomous/utils/markdown.py,sha256=tf8vlHARiQO1X_aGbqlYozzP_TbdiDRT9EEP6aFRQo0,2153
-autonomous_app-0.3.41.dist-info/METADATA,sha256=bF72uTuECc-BhEP0w7uDgm7gqnLsx-cf4eIH1uIxJhY,3024
-autonomous_app-0.3.41.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-autonomous_app-0.3.41.dist-info/top_level.txt,sha256=ZyxWWDdbvZekF3UFunxl4BQsVDb_FOW3eTn0vun_jb4,11
-autonomous_app-0.3.41.dist-info/RECORD,,
+autonomous_app-0.3.42.dist-info/METADATA,sha256=wQyVXrufozI92_4Yv4zfuPya9LYxav2u8jHjUogkQRA,3024
+autonomous_app-0.3.42.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+autonomous_app-0.3.42.dist-info/top_level.txt,sha256=ZyxWWDdbvZekF3UFunxl4BQsVDb_FOW3eTn0vun_jb4,11
+autonomous_app-0.3.42.dist-info/RECORD,,