agentnova 0.3.6__tar.gz → 0.3.7__tar.gz

{agentnova-0.3.6 → agentnova-0.3.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentnova
-Version: 0.3.6
+Version: 0.3.7
 Summary: ⚛️ AgentNova - A minimal, hackable agentic framework for local LLM inference
 Author-email: VTSTech <veritas@vts-tech.org>
 Maintainer-email: VTSTech <veritas@vts-tech.org>
@@ -32,7 +32,7 @@ Requires-Dist: black>=23.0; extra == "dev"
 Requires-Dist: ruff>=0.1.0; extra == "dev"
 Dynamic: license-file
 
-# ⚛️ AgentNova R03.6
+# ⚛️ AgentNova R03.7
 
 **Status: Alpha**
 
@@ -63,7 +63,7 @@ Inspired by the architecture of OpenClaw, rebuilt from scratch for local-first o
 
 - **Zero dependencies** — Uses Python stdlib only (urllib for HTTP)
 - **Ollama + BitNet backends** — Switch with `--backend` flag
-- **Dual API support** — OpenResponses (`--api resp`) and Chat-Completions (`--api comp`)
+- **Dual API support** — OpenResponses (`--api openre`) and OpenAI Chat-Completions (`--api openai`)
 - **Three-tier tool support** — Native, ReAct, or none (auto-detected)
 - **Small model optimized** — Fuzzy matching, argument normalization
 - **Built-in security** — Path validation, command blocklist, SSRF protection
@@ -99,8 +99,8 @@ agentnova chat -m qwen2.5:0.5b --tools calculator,shell
 # Autonomous agent mode
 agentnova agent -m qwen2.5:7b --tools calculator,shell,write_file
 
-# Use Chat-Completions API (OpenAI-compatible)
-agentnova chat -m qwen2.5:0.5b --api comp
+# Use OpenAI Chat-Completions API
+agentnova chat -m qwen2.5:0.5b --api openai
 
 # List available models
 agentnova models
@@ -138,7 +138,7 @@ from agentnova.backends import get_backend
 from agentnova.core.types import ApiMode
 
 # Use Chat-Completions mode with streaming
-backend = get_backend("ollama", api_mode=ApiMode.COMPLETIONS)
+backend = get_backend("ollama", api_mode=ApiMode.OPENAI)
 
 for chunk in backend.generate_completions_stream(
     model="qwen2.5:0.5b",
@@ -261,7 +261,7 @@ agentnova config --urls  # Show only URLs
 
 | Option | Description |
 |--------|-------------|
-| `--api resp\|comp` | API mode: OpenResponses (default) or Chat-Completions |
+| `--api openre\|openai` | API mode: OpenResponses (default) or OpenAI Chat-Completions |
 | `--response-format text\|json` | Response format (Chat-Completions mode) |
 | `--truncation auto\|disabled` | Truncation behavior for long responses |
 | `--soul <path>` | Load Soul Spec persona package |
@@ -283,7 +283,7 @@ agentnova run "What is 2+2?"
 
 ## Tests & Examples
 
-AgentNova includes a suite of tests for validating agent capabilities:
+AgentNova includes a comprehensive suite of tests for validating agent capabilities across reasoning, knowledge, and tool usage:
 
 ```bash
 # Basic agent test (no tools)
@@ -292,7 +292,7 @@ python -m agentnova.examples.00_basic_agent
 # Quick 5-question diagnostic
 python -m agentnova.examples.01_quick_diagnostic
 
-# Tool usage tests (calculator, shell, datetime)
+# Tool usage tests (calculator, shell, datetime, file, python_repl)
 python -m agentnova.examples.02_tool_test
 
 # Logic and reasoning tests (BBH-style)
@@ -300,16 +300,45 @@ python -m agentnova.examples.03_reasoning_test
 
 # GSM8K math benchmark (50 questions)
 python -m agentnova.examples.04_gsm8k_benchmark
+
+# Common sense reasoning (BIG-bench)
+python -m agentnova.examples.05_common_sense
+
+# Causal reasoning (BIG-bench)
+python -m agentnova.examples.06_causal_reasoning
+
+# Logical deduction (BIG-bench)
+python -m agentnova.examples.07_logical_deduction
+
+# Reading comprehension
+python -m agentnova.examples.08_reading_comprehension
+
+# General knowledge (BIG-bench)
+python -m agentnova.examples.09_general_knowledge
+
+# Implicit reasoning
+python -m agentnova.examples.10_implicit_reasoning
+
+# Analogical reasoning
+python -m agentnova.examples.11_analogical_reasoning
 ```
 
 ### Test Categories
 
 | Test | Questions | Focus |
 |------|-----------|-------|
+| Basic Agent | 1 | Single prompt, no tools |
 | Quick Diagnostic | 5 | Calculator tool, multi-step reasoning |
-| Tool Test | 10 | Calculator, shell, datetime tools |
-| Reasoning Test | 13 | Logic, deduction, patterns, spatial |
+| Tool Test | 10 | Calculator, shell, datetime, file, python_repl tools |
+| Reasoning Test | 14 | Logic, deduction, patterns, spatial |
 | GSM8K Benchmark | 50 | Math word problems |
+| Common Sense | 25 | Physical properties, everyday reasoning |
+| Causal Reasoning | 25 | Cause and effect relationships |
+| Logical Deduction | 25 | Formal logic puzzles |
+| Reading Comprehension | 25 | Passage-based Q&A |
+| General Knowledge | 25 | Science, history, geography |
+| Implicit Reasoning | 25 | Unstated assumptions and inference |
+| Analogical Reasoning | 25 | Pattern matching and analogies |
 
 ### Benchmark Results (Quick Diagnostic)
 
@@ -318,8 +347,9 @@ python -m agentnova.examples.04_gsm8k_benchmark
 | functiongemma:270m | 5/5 (100%) | ~20s | native |
 | granite4:350m | 5/5 (100%) | ~50s | native |
 | qwen2.5:0.5b | 5/5 (100%) | 38s | native |
-| qwen2.5-coder:0.5b | 5/5 (100%) | 93s | react |
+| qwen2.5-coder:0.5b | 5/5 (100%) | 93s | native |
 | qwen3:0.6b | 5/5 (100%) | 70s | react |
+| deepseek-r1:1.5b | 5/5 (100%) | ~305s | native |
 
 All tested models achieve 100% on the Quick Diagnostic. Native models are ~2x faster than ReAct models due to direct API tool calling.
 

{agentnova-0.3.6 → agentnova-0.3.7}/README.md

@@ -1,4 +1,4 @@
-# ⚛️ AgentNova R03.6
+# ⚛️ AgentNova R03.7
 
 **Status: Alpha**
 
@@ -29,7 +29,7 @@ Inspired by the architecture of OpenClaw, rebuilt from scratch for local-first o
 
 - **Zero dependencies** — Uses Python stdlib only (urllib for HTTP)
 - **Ollama + BitNet backends** — Switch with `--backend` flag
-- **Dual API support** — OpenResponses (`--api resp`) and Chat-Completions (`--api comp`)
+- **Dual API support** — OpenResponses (`--api openre`) and OpenAI Chat-Completions (`--api openai`)
 - **Three-tier tool support** — Native, ReAct, or none (auto-detected)
 - **Small model optimized** — Fuzzy matching, argument normalization
 - **Built-in security** — Path validation, command blocklist, SSRF protection
@@ -65,8 +65,8 @@ agentnova chat -m qwen2.5:0.5b --tools calculator,shell
 # Autonomous agent mode
 agentnova agent -m qwen2.5:7b --tools calculator,shell,write_file
 
-# Use Chat-Completions API (OpenAI-compatible)
-agentnova chat -m qwen2.5:0.5b --api comp
+# Use OpenAI Chat-Completions API
+agentnova chat -m qwen2.5:0.5b --api openai
 
 # List available models
 agentnova models
@@ -104,7 +104,7 @@ from agentnova.backends import get_backend
 from agentnova.core.types import ApiMode
 
 # Use Chat-Completions mode with streaming
-backend = get_backend("ollama", api_mode=ApiMode.COMPLETIONS)
+backend = get_backend("ollama", api_mode=ApiMode.OPENAI)
 
 for chunk in backend.generate_completions_stream(
     model="qwen2.5:0.5b",
@@ -227,7 +227,7 @@ agentnova config --urls  # Show only URLs
 
 | Option | Description |
 |--------|-------------|
-| `--api resp\|comp` | API mode: OpenResponses (default) or Chat-Completions |
+| `--api openre\|openai` | API mode: OpenResponses (default) or OpenAI Chat-Completions |
 | `--response-format text\|json` | Response format (Chat-Completions mode) |
 | `--truncation auto\|disabled` | Truncation behavior for long responses |
 | `--soul <path>` | Load Soul Spec persona package |
@@ -249,7 +249,7 @@ agentnova run "What is 2+2?"
 
 ## Tests & Examples
 
-AgentNova includes a suite of tests for validating agent capabilities:
+AgentNova includes a comprehensive suite of tests for validating agent capabilities across reasoning, knowledge, and tool usage:
 
 ```bash
 # Basic agent test (no tools)
@@ -258,7 +258,7 @@ python -m agentnova.examples.00_basic_agent
 # Quick 5-question diagnostic
 python -m agentnova.examples.01_quick_diagnostic
 
-# Tool usage tests (calculator, shell, datetime)
+# Tool usage tests (calculator, shell, datetime, file, python_repl)
 python -m agentnova.examples.02_tool_test
 
 # Logic and reasoning tests (BBH-style)
@@ -266,16 +266,45 @@ python -m agentnova.examples.03_reasoning_test
 
 # GSM8K math benchmark (50 questions)
 python -m agentnova.examples.04_gsm8k_benchmark
+
+# Common sense reasoning (BIG-bench)
+python -m agentnova.examples.05_common_sense
+
+# Causal reasoning (BIG-bench)
+python -m agentnova.examples.06_causal_reasoning
+
+# Logical deduction (BIG-bench)
+python -m agentnova.examples.07_logical_deduction
+
+# Reading comprehension
+python -m agentnova.examples.08_reading_comprehension
+
+# General knowledge (BIG-bench)
+python -m agentnova.examples.09_general_knowledge
+
+# Implicit reasoning
+python -m agentnova.examples.10_implicit_reasoning
+
+# Analogical reasoning
+python -m agentnova.examples.11_analogical_reasoning
 ```
 
 ### Test Categories
 
 | Test | Questions | Focus |
 |------|-----------|-------|
+| Basic Agent | 1 | Single prompt, no tools |
 | Quick Diagnostic | 5 | Calculator tool, multi-step reasoning |
-| Tool Test | 10 | Calculator, shell, datetime tools |
-| Reasoning Test | 13 | Logic, deduction, patterns, spatial |
+| Tool Test | 10 | Calculator, shell, datetime, file, python_repl tools |
+| Reasoning Test | 14 | Logic, deduction, patterns, spatial |
 | GSM8K Benchmark | 50 | Math word problems |
+| Common Sense | 25 | Physical properties, everyday reasoning |
+| Causal Reasoning | 25 | Cause and effect relationships |
+| Logical Deduction | 25 | Formal logic puzzles |
+| Reading Comprehension | 25 | Passage-based Q&A |
+| General Knowledge | 25 | Science, history, geography |
+| Implicit Reasoning | 25 | Unstated assumptions and inference |
+| Analogical Reasoning | 25 | Pattern matching and analogies |
 
 ### Benchmark Results (Quick Diagnostic)
 
@@ -284,8 +313,9 @@ python -m agentnova.examples.04_gsm8k_benchmark
 | functiongemma:270m | 5/5 (100%) | ~20s | native |
 | granite4:350m | 5/5 (100%) | ~50s | native |
 | qwen2.5:0.5b | 5/5 (100%) | 38s | native |
-| qwen2.5-coder:0.5b | 5/5 (100%) | 93s | react |
+| qwen2.5-coder:0.5b | 5/5 (100%) | 93s | native |
 | qwen3:0.6b | 5/5 (100%) | 70s | react |
+| deepseek-r1:1.5b | 5/5 (100%) | ~305s | native |
 
 All tested models achieve 100% on the Quick Diagnostic. Native models are ~2x faster than ReAct models due to direct API tool calling.
 
@@ -319,4 +349,4 @@ Contributions welcome! Please read the contributing guidelines first.
 
 - Built for local inference with [Ollama](https://ollama.ai)
 - Optimized for small, efficient models
-- Inspired by ReAct and other agentic frameworks
+- Inspired by ReAct and other agentic frameworks

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/__init__.py

@@ -28,7 +28,7 @@ Example Usage:
     agent = Agent(model="qwen2.5:0.5b", soul="/path/to/soul/package")
 """
 
-__version__ = "0.3.6"
+__version__ = "0.3.7"
 __author__ = "VTSTech"
 __status__ = "Alpha"
 

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/agent.py

@@ -314,7 +314,7 @@ class Agent:
     @property
     def _is_comp_mode(self) -> bool:
         """Check if backend is using Chat-Completions (comp) API mode."""
-        return hasattr(self.backend, 'api_mode') and self.backend.api_mode == ApiMode.COMPLETIONS
+        return hasattr(self.backend, 'api_mode') and self.backend.api_mode == ApiMode.OPENAI
 
     def _log_openresponses(self, msg: str) -> None:
         """Log OpenResponses debug message only when not in comp mode."""

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/backends/__init__.py

@@ -26,7 +26,7 @@ def get_backend(name: str, timeout: int | None = None, api_mode: ApiMode | str |
     Args:
         name: Backend name ("ollama", "bitnet")
         timeout: Request timeout in seconds (default: 120)
-        api_mode: API mode ("resp" for OpenResponses/native, "comp" for Chat-Completions)
+        api_mode: API mode ("openre" for OpenResponses, "openai" for Chat-Completions)
         **kwargs: Backend-specific configuration
 
     Returns:
@@ -59,7 +59,7 @@ def get_backend(name: str, timeout: int | None = None, api_mode: ApiMode | str |
     return backend_class(**kwargs)
 
 
-def get_default_backend(name: str | None = None, api_mode: ApiMode | str | None = None) -> BaseBackend:
+def get_default_backend(name: str | None = None, api_mode: ApiMode | str | None = None, timeout: int | None = None) -> BaseBackend:
     """
     Get a default backend instance.
 
@@ -67,13 +67,14 @@ def get_default_backend(name: str | None = None, api_mode: ApiMode | str | None
 
     Args:
         name: Backend name (optional, uses env var if not provided)
-        api_mode: API mode ("resp" for OpenResponses/native, "comp" for Chat-Completions)
+        api_mode: API mode ("openre" for OpenResponses/native, "openai" for Chat-Completions)
+        timeout: Request timeout in seconds (default: 120)
 
     Returns:
         Backend instance with default configuration
     """
     backend_name = name or AGENTNOVA_BACKEND
-    return get_backend(backend_name, api_mode=api_mode)
+    return get_backend(backend_name, api_mode=api_mode, timeout=timeout)
 
 
 def register_backend(name: str, backend_class: type[BaseBackend]) -> None:

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/backends/ollama.py

@@ -32,7 +32,7 @@ class OllamaBackend(BaseBackend):
         host: str | None = None,
         port: int | None = None,
         config: BackendConfig | None = None,
-        api_mode: ApiMode | str = ApiMode.RESPONSES,
+        api_mode: ApiMode | str = ApiMode.OPENRE,
     ):
         # Determine base URL - priority: base_url > host/port > env > default
         if base_url:
@@ -47,7 +47,7 @@ class OllamaBackend(BaseBackend):
         else:
             super().__init__(BackendConfig())
         
-        # Set API mode (resp = OpenResponses/native, comp = Chat-Completions)
+        # Set API mode (openre = OpenResponses, openai = Chat-Completions)
         if isinstance(api_mode, str):
             api_mode = ApiMode(api_mode.lower())
         self._api_mode = api_mode
@@ -137,7 +137,7 @@ class OllamaBackend(BaseBackend):
             **kwargs: Additional options passed to Ollama
         """
         # Dispatch based on api_mode
-        if self._api_mode == ApiMode.COMPLETIONS:
+        if self._api_mode == ApiMode.OPENAI:
             if os.environ.get("AGENTNOVA_DEBUG"):
                 print(f"  [Ollama] Dispatching to OpenAI-compatible API (mode={self._api_mode.value})")
             return self.generate_completions(

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/cli.py

@@ -171,8 +171,8 @@ def create_parser() -> argparse.ArgumentParser:
     run_parser.add_argument("-m", "--model", default=None, help="Model to use")
     run_parser.add_argument("--tools", default="calculator", help="Comma-separated tool list")
     run_parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None, help="Backend to use")
-    run_parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                           help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    run_parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                           help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions)")
     run_parser.add_argument("--stream", action="store_true", help="Stream output")
     run_parser.add_argument("--debug", action="store_true", help="Enable debug output")
     run_parser.add_argument("-v", "--verbose", action="store_true", help="Verbose output")
@@ -202,8 +202,8 @@ def create_parser() -> argparse.ArgumentParser:
     chat_parser.add_argument("-m", "--model", default=None, help="Model to use")
     chat_parser.add_argument("--tools", default="", help="Comma-separated tool list")
     chat_parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None, help="Backend to use")
-    chat_parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                           help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    chat_parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                           help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions)")
     chat_parser.add_argument("--debug", action="store_true", help="Enable debug output")
     chat_parser.add_argument("--force-react", action="store_true", help="Force ReAct mode")
     chat_parser.add_argument("--soul", default=None, help="Path to Soul Spec package (disabled by default)")
@@ -231,8 +231,8 @@ def create_parser() -> argparse.ArgumentParser:
     agent_parser.add_argument("-m", "--model", default=None, help="Model to use")
     agent_parser.add_argument("--tools", default="calculator,shell,write_file", help="Comma-separated tool list")
     agent_parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None, help="Backend to use")
-    agent_parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                           help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    agent_parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                           help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions)")
     agent_parser.add_argument("--debug", action="store_true", help="Enable debug output")
     agent_parser.add_argument("--force-react", action="store_true", help="Force ReAct mode for tool calling")
     agent_parser.add_argument("--soul", default=None, help="Path to Soul Spec package (disabled by default)")
@@ -271,8 +271,8 @@ def create_parser() -> argparse.ArgumentParser:
     test_parser.add_argument("-m", "--model", default=None, 
                              help="Model to test (supports patterns: 'qwen', 'g', ':0.5b')")
     test_parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None, help="Backend to use")
-    test_parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                           help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    test_parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                           help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions)")
     test_parser.add_argument("--debug", action="store_true", help="Enable debug output")
     test_parser.add_argument("--list", action="store_true", help="List available tests")
     test_parser.add_argument("--acp", action="store_true", help="Enable ACP logging to Agent Control Panel")
@@ -295,6 +295,10 @@ def create_parser() -> argparse.ArgumentParser:
     test_parser.add_argument("--soul", default=None, help="Path to Soul Spec package (disabled by default)")
     test_parser.add_argument("--soul-level", type=int, default=2, choices=[1, 2, 3],
                            help="Soul progressive disclosure level (1=quick, 2=full, 3=deep)")
+    test_parser.add_argument("--tools-only", action="store_true", dest="tools_only",
+                           help="Only run Phase 1 (direct tool tests, no model)")
+    test_parser.add_argument("--model-only", action="store_true", dest="model_only",
+                           help="Only run Phase 2 (model tool calling tests)")
 
     # Version command
     subparsers.add_parser("version", help="Show version information")
@@ -360,7 +364,7 @@ def cmd_run(args: argparse.Namespace) -> int:
     config = get_config()
     model = args.model or config.default_model
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
+    api_mode = getattr(args, 'api_mode', 'openre')
 
     # Initialize ACP if requested
     acp, should_stop = _init_acp(args, config, "AgentNova-Run")
@@ -386,7 +390,7 @@ def cmd_run(args: argparse.Namespace) -> int:
         force_react=getattr(args, 'force_react', False),
         soul=getattr(args, 'soul', None),
         soul_level=getattr(args, 'soul_level', 2),
-        num_ctx=getattr(args, 'num_ctx', None) or config.num_ctx,
+        num_ctx=getattr(args, 'num_ctx', None) if getattr(args, 'num_ctx', None) is not None else config.num_ctx,
         temperature=getattr(args, 'temperature', None),
         top_p=getattr(args, 'top_p', None),
         num_predict=getattr(args, 'num_predict', None),
@@ -414,7 +418,7 @@ def cmd_chat(args: argparse.Namespace) -> int:
     config = get_config()
     model = args.model or config.default_model
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
+    api_mode = getattr(args, 'api_mode', 'openre')
 
     # Initialize ACP if requested
     acp, should_stop = _init_acp(args, config, "AgentNova-Chat")
@@ -439,7 +443,7 @@ def cmd_chat(args: argparse.Namespace) -> int:
         debug=args.debug,
         soul=getattr(args, 'soul', None),
         soul_level=getattr(args, 'soul_level', 2),
-        num_ctx=getattr(args, 'num_ctx', None) or config.num_ctx,
+        num_ctx=getattr(args, 'num_ctx', None) if getattr(args, 'num_ctx', None) is not None else config.num_ctx,
         temperature=getattr(args, 'temperature', None),
         top_p=getattr(args, 'top_p', None),
         num_predict=getattr(args, 'num_predict', None),
@@ -512,7 +516,7 @@ def cmd_agent(args: argparse.Namespace) -> int:
     config = get_config()
     model = args.model or config.default_model
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
+    api_mode = getattr(args, 'api_mode', 'openre')
 
     # Initialize ACP if requested
     acp, should_stop = _init_acp(args, config, "AgentNova-Agent")
@@ -537,7 +541,7 @@ def cmd_agent(args: argparse.Namespace) -> int:
         force_react=getattr(args, 'force_react', False),
         soul=getattr(args, 'soul', None),
         soul_level=getattr(args, 'soul_level', 2),
-        num_ctx=getattr(args, 'num_ctx', None) or config.num_ctx,
+        num_ctx=getattr(args, 'num_ctx', None) if getattr(args, 'num_ctx', None) is not None else config.num_ctx,
         temperature=getattr(args, 'temperature', None),
         top_p=getattr(args, 'top_p', None),
         num_predict=getattr(args, 'num_predict', None),
@@ -950,7 +954,7 @@ def cmd_test(args: argparse.Namespace) -> int:
     # Check backend
     config = get_config()
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
+    api_mode = getattr(args, 'api_mode', 'openre')
     timeout = getattr(args, 'timeout', None)
     backend = get_backend(backend_name, timeout=timeout, api_mode=api_mode)
     
@@ -1014,7 +1018,7 @@ def cmd_test(args: argparse.Namespace) -> int:
         print(f"{dim('Model:')} {cyan(models_to_test[0])}")
     else:
         print(f"{dim('Models:')} {cyan(str(len(models_to_test)))} matching '{model_pattern}'")
-    num_ctx_val = getattr(args, 'num_ctx', None) or config.num_ctx
+    num_ctx_val = getattr(args, 'num_ctx', None) if getattr(args, 'num_ctx', None) is not None else config.num_ctx
     if num_ctx_val:
         ctx_display = f"{num_ctx_val // 1024}K" if num_ctx_val >= 1024 else str(num_ctx_val)
         print(f"{dim('Context:')} {yellow(ctx_display)}")
@@ -1056,7 +1060,7 @@ def cmd_test(args: argparse.Namespace) -> int:
                     test_argv.append("--debug")
                 if args.backend:
                     test_argv.extend(["--backend", args.backend])
-                if getattr(args, 'api_mode', 'resp') != 'resp':
+                if getattr(args, 'api_mode', 'openre') != 'openre':
                     test_argv.extend(["--api", args.api_mode])
                 if getattr(args, 'force_react', False):
                     test_argv.append("--force-react")
@@ -1077,6 +1081,10 @@ def cmd_test(args: argparse.Namespace) -> int:
                     test_argv.extend(["--temp", str(args.temperature)])
                 if getattr(args, 'top_p', None) is not None:
                     test_argv.extend(["--top-p", str(args.top_p)])
+                if getattr(args, 'tools_only', False):
+                    test_argv.append("--tools-only")
+                if getattr(args, 'model_only', False):
+                    test_argv.append("--model-only")
                 
                 # Override sys.argv for the test module's argparse
                 old_argv = sys.argv

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/core/types.py

@@ -74,15 +74,15 @@ class BackendType(Enum):
 class ApiMode(Enum):
     """API mode for backend communication.
     
-    - RESPONSES: OpenResponses-style API (Ollama native /api/chat)
-    - COMPLETIONS: OpenAI Chat-Completions compatible API (/v1/chat/completions)
+    - OPENRE: OpenResponses API (open spec for agentic workflows)
+    - OPENAI: OpenAI Chat-Completions API
     
     Ollama supports both endpoints:
-    - /api/chat - Native Ollama format (default)
+    - /api/chat - OpenResponses format (default)
     - /v1/chat/completions - OpenAI-compatible format
     """
-    RESPONSES = "resp"      # OpenResponses / Ollama native format
-    COMPLETIONS = "comp"    # OpenAI Chat-Completions compatible format
+    OPENRE = "openre"      # OpenResponses API (/api/chat)
+    OPENAI = "openai"      # OpenAI Chat-Completions API (/v1/chat/completions)
 
 
 # Type aliases for clarity

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/examples/00_basic_agent.py

@@ -28,8 +28,22 @@ def parse_args():
     parser.add_argument("-m", "--model", default=None, help="Model to use")
     parser.add_argument("--debug", action="store_true", help="Enable debug output")
     parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None)
-    parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                       help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                       help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions (OpenAI))")
+    parser.add_argument("--soul", default=None, help="Path to Soul Spec package")
+    parser.add_argument("--soul-level", type=int, default=2, choices=[1, 2, 3],
+                       help="Soul progressive disclosure level")
+    parser.add_argument("--num-ctx", type=int, default=None,
+                       help="Context window size in tokens")
+    parser.add_argument("--num-predict", type=int, default=None,
+                       help="Maximum tokens to generate")
+    parser.add_argument("--temp", type=float, default=None, dest="temperature",
+                       help="Sampling temperature 0.0-2.0")
+    parser.add_argument("--top-p", type=float, default=None, dest="top_p",
+                       help="Nucleus sampling probability 0.0-1.0")
+    parser.add_argument("--force-react", action="store_true", help="Force ReAct mode for tool calling")
+    parser.add_argument("--timeout", type=int, default=None,
+                       help="Request timeout in seconds (default: 120)")
     return parser.parse_args()
 
 
@@ -39,8 +53,9 @@ def main():
     
     # Get backend (respects AGENTNOVA_BACKEND env var)
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
-    backend = get_default_backend(backend_name, api_mode=api_mode)
+    api_mode = getattr(args, 'api_mode', 'openre')
+    timeout = getattr(args, 'timeout', None)
+    backend = get_default_backend(backend_name, api_mode=api_mode, timeout=timeout)
     
     # Check if backend is running
     if not backend.is_running():
@@ -57,13 +72,22 @@ def main():
         model=args.model or config.default_model,
         backend=backend,
         debug=args.debug or config.debug,
+        soul=args.soul,
+        soul_level=args.soul_level,
+        force_react=getattr(args, 'force_react', False),
+        num_ctx=getattr(args, 'num_ctx', None),
+        num_predict=getattr(args, 'num_predict', None),
+        temperature=getattr(args, 'temperature', None),
+        top_p=getattr(args, 'top_p', None),
     )
     
     print(f"\n⚛️ AgentNova Basic Agent")
     print(f"   Model: {args.model or config.default_model}")
     print(f"   Backend: {backend_name}")
-    if api_mode != 'resp':
+    if api_mode != 'openre':
         print(f"   API Mode: {api_mode}")
+    if timeout:
+        print(f"   Timeout: {timeout}s")
     print("-" * 40)
     
     # Single question

{agentnova-0.3.6 → agentnova-0.3.7}/agentnova/examples/01_quick_diagnostic.py

@@ -43,8 +43,8 @@ def parse_args():
     parser.add_argument("-m", "--model", default=None, help="Model to test")
     parser.add_argument("--debug", action="store_true", help="Enable debug output")
     parser.add_argument("--backend", choices=["ollama", "bitnet"], default=None)
-    parser.add_argument("--api", choices=["resp", "comp"], default="resp", dest="api_mode",
-                       help="API mode: 'resp' (OpenResponses/native) or 'comp' (Chat-Completions)")
+    parser.add_argument("--api", choices=["openre", "openai"], default="openre", dest="api_mode",
+                       help="API mode: 'openre' (OpenResponses) or 'openai' (Chat-Completions (OpenAI))")
     parser.add_argument("--force-react", action="store_true", help="Force ReAct mode for tool calling")
     parser.add_argument("--soul", default=None, help="Path to Soul Spec package")
     parser.add_argument("--soul-level", type=int, default=2, choices=[1, 2, 3],
@@ -222,7 +222,7 @@ def main():
     
     # Get backend with timeout
     backend_name = args.backend or config.backend
-    api_mode = getattr(args, 'api_mode', 'resp')
+    api_mode = getattr(args, 'api_mode', 'openre')
     timeout = getattr(args, 'timeout', None)
     backend = get_backend(backend_name, timeout=timeout, api_mode=api_mode)
     
@@ -242,8 +242,8 @@ def main():
     print(f"   Backend: {backend_name} ({backend.base_url})")
     print(f"   Model: {model}")
     api_mode_display = {
-        'resp': '[OpenAI] OpenResponses (2025)',
-        'comp': '[OpenAI] ChatCompletions (2023)'
+        'openre': '[OpenResponses]',
+        'openai': '[OpenAI] ChatCompletions'
     }.get(api_mode, api_mode)
     print(f"   API Mode: {api_mode_display}")
     if timeout:
@@ -252,7 +252,9 @@ def main():
         print(f"   Force ReAct: True")
     if args.soul:
         print(f"   Soul: {args.soul}")
-    num_ctx = getattr(args, 'num_ctx', None) or getattr(config, 'num_ctx', None)
+    num_ctx = getattr(args, 'num_ctx', None)
+    if num_ctx is None:
+        num_ctx = getattr(config, 'num_ctx', None)
     if num_ctx:
         ctx_display = f"{num_ctx // 1024}K" if num_ctx >= 1024 else str(num_ctx)
         print(f"   Context: {ctx_display}")
@@ -274,5 +276,4 @@ def main():
 
 
 if __name__ == "__main__":
-    result = main()
-    sys.exit(result.get("exit_code", 0))
+    result = main()