refactor: centralize model IDs into model_config.py

libs/openant-core/context/application_context.py

@@ -32,6 +32,14 @@
 from anthropic import Anthropic
 from dotenv import load_dotenv
 
+# Ensure libs/openant-core is on sys.path so `utilities.*` imports resolve
+# regardless of how this module is loaded.
+_OPENANT_CORE_ROOT = str(Path(__file__).parent.parent)
+if _OPENANT_CORE_ROOT not in sys.path:
+    sys.path.insert(0, _OPENANT_CORE_ROOT)
+
+from utilities.model_config import MODEL_AUXILIARY  # noqa: E402
+
 # Load environment variables
 load_dotenv()
 
@@ -462,7 +470,7 @@ def _build_type_descriptions() -> str:
 
 def generate_application_context(
     repo_path: Path,
-    model: str = "claude-sonnet-4-20250514",
+    model: str = MODEL_AUXILIARY,
     force_regenerate: bool = False,
 ) -> ApplicationContext:
     """Generate application context using LLM analysis.

libs/openant-core/context/generate_context.py

@@ -14,6 +14,7 @@
 # Add parent directory to path for imports
 sys.path.insert(0, str(Path(__file__).parent.parent))
 
+from utilities.model_config import MODEL_AUXILIARY
 from context.application_context import (
     ApplicationType,
     APPLICATION_TYPE_INFO,
@@ -78,8 +79,8 @@ def main():
 
     parser.add_argument(
         "--model", "-m",
-        default="claude-sonnet-4-20250514",
-        help="Anthropic model to use (default: claude-sonnet-4-20250514)",
+        default=MODEL_AUXILIARY,
+        help=f"Anthropic model to use (default: {MODEL_AUXILIARY})",
     )
 
     parser.add_argument(

libs/openant-core/core/analyzer.py

@@ -313,7 +313,8 @@ def run_analysis(
     checkpoint.dir = checkpoint_path
 
     # Select model
-    model_id = "claude-opus-4-6" if model == "opus" else "claude-sonnet-4-20250514"
+    from utilities.model_config import MODEL_AUXILIARY, MODEL_PRIMARY
+    model_id = MODEL_PRIMARY if model == "opus" else MODEL_AUXILIARY
     print(f"[Analyze] Model: {model_id}", file=sys.stderr)
 
     # Initialize client

libs/openant-core/core/enhancer.py

@@ -50,7 +50,8 @@ def enhance_dataset(
     # Configure global rate limiter
     configure_rate_limiter(backoff_seconds=float(backoff_seconds))
 
-    model_id = "claude-sonnet-4-20250514" if model == "sonnet" else "claude-opus-4-6"
+    from utilities.model_config import MODEL_AUXILIARY, MODEL_PRIMARY
+    model_id = MODEL_AUXILIARY if model == "sonnet" else MODEL_PRIMARY
     print(f"[Enhance] Mode: {mode}", file=sys.stderr)
     print(f"[Enhance] Model: {model_id}", file=sys.stderr)
 

libs/openant-core/core/reporter.py

@@ -587,11 +587,12 @@ def _record_usage_in_tracker(usage: dict):
     """Record usage in the global TokenTracker so step_context captures it."""
     try:
         from utilities.llm_client import get_global_tracker
+        from utilities.model_config import MODEL_PRIMARY
         tracker = get_global_tracker()
         # Record as a single aggregated call
         if usage.get("total_tokens", 0) > 0:
             tracker.record_call(
-                model="claude-opus-4-6",
+                model=MODEL_PRIMARY,
                 input_tokens=usage["input_tokens"],
                 output_tokens=usage["output_tokens"],
             )

libs/openant-core/experiment.py

@@ -474,7 +474,8 @@ def run_experiment(
         Experiment results with metrics
     """
     # Select model
-    model_id = "claude-opus-4-20250514" if model == "opus" else "claude-sonnet-4-20250514"
+    from utilities.model_config import MODEL_AUXILIARY, MODEL_PRIMARY
+    model_id = MODEL_PRIMARY if model == "opus" else MODEL_AUXILIARY
     print(f"Using model: {model_id}")
     print(f"Enhanced context: {enhanced}")
     print(f"Context correction: {correct_context}")

libs/openant-core/generate_report.py

@@ -27,16 +27,26 @@
 import json
 import html
 import os
+import sys
 from datetime import datetime
+from pathlib import Path
 
 import anthropic
 from dotenv import load_dotenv
 
+# Ensure libs/openant-core is on sys.path so `utilities.*` imports resolve
+# regardless of the caller's working directory.
+_OPENANT_CORE_ROOT = str(Path(__file__).parent)
+if _OPENANT_CORE_ROOT not in sys.path:
+    sys.path.insert(0, _OPENANT_CORE_ROOT)
+
+from utilities.model_config import MODEL_AUXILIARY  # noqa: E402
+
 # Load environment variables from .env file
 load_dotenv()
 
 
-REPORT_MODEL = "claude-sonnet-4-20250514"
+REPORT_MODEL = MODEL_AUXILIARY
 MAX_TOKENS = 4096
 
 

libs/openant-core/openant/cli.py

@@ -810,9 +810,10 @@ def cmd_report_data(args):
 {findings_text}
 """
                 print("[Report] Generating remediation guidance (LLM)...", file=sys.stderr)
+                from utilities.model_config import MODEL_AUXILIARY
                 client = anthropic.Anthropic()
                 response = client.messages.create(
-                    model="claude-sonnet-4-20250514",
+                    model=MODEL_AUXILIARY,
                     max_tokens=4096,
                     messages=[{"role": "user", "content": prompt}],
                 )
@@ -829,7 +830,7 @@ def _linkify_finding(m):
                 usage = response.usage
                 tracker = get_global_tracker()
                 tracker.record_call(
-                    model="claude-sonnet-4-20250514",
+                    model=MODEL_AUXILIARY,
                     input_tokens=usage.input_tokens,
                     output_tokens=usage.output_tokens,
                 )

libs/openant-core/report/generator.py

@@ -12,18 +12,25 @@
 from pathlib import Path
 from dotenv import load_dotenv
 
+# Ensure libs/openant-core is on sys.path so `utilities.*` imports resolve.
+_OPENANT_CORE_ROOT = str(Path(__file__).parent.parent)
+if _OPENANT_CORE_ROOT not in sys.path:
+    sys.path.insert(0, _OPENANT_CORE_ROOT)
+
+from utilities.model_config import MODEL_PRIMARY, MODEL_AUXILIARY  # noqa: E402
+
 from .schema import validate_pipeline_output, ValidationError
 
 load_dotenv()
 
 PROMPTS_DIR = Path(__file__).parent / "prompts"
-MODEL = "claude-opus-4-6"
+MODEL = MODEL_PRIMARY
 
-# Pricing per million tokens
+# Pricing per million tokens. Aliased opus keys retained for backward
+# compatibility with any cached/legacy responses that recorded the literal.
 _PRICING = {
-    "claude-opus-4-6": {"input": 15.00, "output": 75.00},
-    "claude-opus-4-20250514": {"input": 15.00, "output": 75.00},
-    "claude-sonnet-4-20250514": {"input": 3.00, "output": 15.00},
+    MODEL_PRIMARY: {"input": 15.00, "output": 75.00},
+    MODEL_AUXILIARY: {"input": 3.00, "output": 15.00},
 }
 _DEFAULT_PRICING = {"input": 3.00, "output": 15.00}
 

libs/openant-core/tests/test_model_config.py

@@ -0,0 +1,93 @@
+"""Tests for the central model configuration module."""
+import re
+from pathlib import Path
+
+import pytest
+
+from utilities import model_config
+from utilities.model_config import MODEL_AUXILIARY, MODEL_DEFAULT, MODEL_PRIMARY
+
+
+# Regex for a valid Claude model identifier (e.g. claude-opus-4-20250514,
+# claude-sonnet-4-6, claude-haiku-4-5).
+_MODEL_ID_RE = re.compile(r"^claude-(opus|sonnet|haiku)-[0-9A-Za-z-]+$")
+
+# Regex used by the regression test to detect any hardcoded
+# claude-opus-* / claude-sonnet-* string literal.
+_HARDCODED_LITERAL_RE = re.compile(r"claude-(?:opus|sonnet)-[0-9][0-9A-Za-z-]*")
+
+
+class TestModelConstants:
+    """Constants must exist, be non-empty strings, and match Claude model id format."""
+
+    def test_model_primary_is_valid_string(self):
+        assert isinstance(MODEL_PRIMARY, str)
+        assert MODEL_PRIMARY, "MODEL_PRIMARY must be non-empty"
+        assert _MODEL_ID_RE.match(MODEL_PRIMARY), (
+            f"MODEL_PRIMARY={MODEL_PRIMARY!r} does not match expected "
+            f"claude-(opus|sonnet|haiku)-... format"
+        )
+
+    def test_model_auxiliary_is_valid_string(self):
+        assert isinstance(MODEL_AUXILIARY, str)
+        assert MODEL_AUXILIARY, "MODEL_AUXILIARY must be non-empty"
+        assert _MODEL_ID_RE.match(MODEL_AUXILIARY), (
+            f"MODEL_AUXILIARY={MODEL_AUXILIARY!r} does not match expected "
+            f"claude-(opus|sonnet|haiku)-... format"
+        )
+
+    def test_model_default_is_valid_string(self):
+        assert isinstance(MODEL_DEFAULT, str)
+        assert MODEL_DEFAULT, "MODEL_DEFAULT must be non-empty"
+        assert _MODEL_ID_RE.match(MODEL_DEFAULT)
+
+    def test_module_exposes_all_three_constants(self):
+        for name in ("MODEL_PRIMARY", "MODEL_AUXILIARY", "MODEL_DEFAULT"):
+            assert hasattr(model_config, name), f"model_config missing {name}"
+
+
+class TestNoHardcodedModelLiterals:
+    """Regression test: no hardcoded claude-opus-*/claude-sonnet-* literals
+    may reappear in libs/openant-core/*.py outside of model_config.py.
+
+    If this test fails, replace the offending literal with an import of
+    MODEL_PRIMARY / MODEL_AUXILIARY / MODEL_DEFAULT from utilities.model_config.
+    """
+
+    # Path to libs/openant-core (this file is at libs/openant-core/tests/...)
+    _CORE_ROOT = Path(__file__).parent.parent
+
+    # Files exempt from the scan (the constants live here, by design)
+    _EXEMPT = {
+        _CORE_ROOT / "utilities" / "model_config.py",
+        # The regression test itself contains the regex pattern source.
+        Path(__file__).resolve(),
+    }
+
+    def test_no_hardcoded_model_strings_outside_model_config(self):
+        offenders: list[tuple[Path, int, str]] = []
+
+        for py_path in self._CORE_ROOT.rglob("*.py"):
+            resolved = py_path.resolve()
+            if resolved in {p.resolve() for p in self._EXEMPT}:
+                continue
+
+            try:
+                text = py_path.read_text(encoding="utf-8")
+            except (OSError, UnicodeDecodeError):
+                continue
+
+            for lineno, line in enumerate(text.splitlines(), start=1):
+                if _HARDCODED_LITERAL_RE.search(line):
+                    offenders.append((py_path, lineno, line.strip()))
+
+        if offenders:
+            details = "\n".join(
+                f"  {path.relative_to(self._CORE_ROOT)}:{lineno}: {snippet}"
+                for path, lineno, snippet in offenders
+            )
+            pytest.fail(
+                "Found hardcoded claude-opus-*/claude-sonnet-* literals outside "
+                "utilities/model_config.py. Replace them with imports from "
+                "utilities.model_config:\n" + details
+            )

libs/openant-core/tests/test_token_tracker.py

@@ -1,5 +1,6 @@
 """Tests for TokenTracker."""
 from utilities.llm_client import TokenTracker, MODEL_PRICING
+from utilities.model_config import MODEL_PRIMARY, MODEL_AUXILIARY
 
 
 class TestTokenTracker:
@@ -13,9 +14,9 @@ def test_initial_state(self):
 
     def test_record_call_known_model(self):
         tracker = TokenTracker()
-        result = tracker.record_call("claude-sonnet-4-20250514", 1000, 500)
+        result = tracker.record_call(MODEL_AUXILIARY, 1000, 500)
 
-        assert result["model"] == "claude-sonnet-4-20250514"
+        assert result["model"] == MODEL_AUXILIARY
         assert result["input_tokens"] == 1000
         assert result["output_tokens"] == 500
         # Sonnet: $3/M input, $15/M output
@@ -31,8 +32,8 @@ def test_record_call_unknown_model_uses_default(self):
 
     def test_cumulative_tracking(self):
         tracker = TokenTracker()
-        tracker.record_call("claude-sonnet-4-20250514", 1000, 500)
-        tracker.record_call("claude-sonnet-4-20250514", 2000, 1000)
+        tracker.record_call(MODEL_AUXILIARY, 1000, 500)
+        tracker.record_call(MODEL_AUXILIARY, 2000, 1000)
 
         assert tracker.total_input_tokens == 3000
         assert tracker.total_output_tokens == 1500
@@ -41,7 +42,7 @@ def test_cumulative_tracking(self):
 
     def test_reset(self):
         tracker = TokenTracker()
-        tracker.record_call("claude-sonnet-4-20250514", 1000, 500)
+        tracker.record_call(MODEL_AUXILIARY, 1000, 500)
         tracker.reset()
 
         assert tracker.total_input_tokens == 0
@@ -51,7 +52,7 @@ def test_reset(self):
 
     def test_get_summary_includes_calls(self):
         tracker = TokenTracker()
-        tracker.record_call("claude-sonnet-4-20250514", 100, 50)
+        tracker.record_call(MODEL_AUXILIARY, 100, 50)
         summary = tracker.get_summary()
 
         assert summary["total_calls"] == 1
@@ -60,14 +61,14 @@ def test_get_summary_includes_calls(self):
 
     def test_get_totals_excludes_calls(self):
         tracker = TokenTracker()
-        tracker.record_call("claude-sonnet-4-20250514", 100, 50)
+        tracker.record_call(MODEL_AUXILIARY, 100, 50)
         totals = tracker.get_totals()
 
         assert totals["total_calls"] == 1
         assert "calls" not in totals
 
     def test_opus_pricing(self):
         tracker = TokenTracker()
-        result = tracker.record_call("claude-opus-4-20250514", 1_000_000, 1_000_000)
+        result = tracker.record_call(MODEL_PRIMARY, 1_000_000, 1_000_000)
         # Opus: $15/M input, $75/M output
         assert result["cost_usd"] == 90.0

libs/openant-core/utilities/agentic_enhancer/agent.py

@@ -17,6 +17,7 @@
 import anthropic
 
 from ..llm_client import TokenTracker, get_global_tracker
+from ..model_config import MODEL_AUXILIARY
 from ..rate_limiter import get_rate_limiter
 from .repository_index import RepositoryIndex
 from .tools import TOOL_DEFINITIONS, ToolExecutor
@@ -26,7 +27,7 @@
 
 
 # Use Sonnet for exploration (cost-effective)
-AGENT_MODEL = "claude-sonnet-4-20250514"
+AGENT_MODEL = MODEL_AUXILIARY
 
 # Safety limits
 MAX_ITERATIONS = 20

libs/openant-core/utilities/context_corrector.py

@@ -17,6 +17,7 @@
 from typing import Optional
 
 from .llm_client import AnthropicClient, TokenTracker, get_global_tracker
+from .model_config import MODEL_AUXILIARY
 
 
 # Maximum characters per batch (leaving room for prompt overhead)
@@ -102,7 +103,7 @@ def parse_missing_context_with_llm(
     prompt = get_missing_context_prompt(reasoning)
 
     try:
-        llm_response = client.analyze_sync(prompt, model="claude-sonnet-4-20250514")
+        llm_response = client.analyze_sync(prompt, model=MODEL_AUXILIARY)
         parsed = _parse_json_response(llm_response)
 
         if parsed and "missing_context" in parsed:
@@ -254,7 +255,7 @@ def search_files_for_context(
         prompt = get_file_search_prompt(missing_context, files_content, batch_info)
 
         try:
-            response = client.analyze_sync(prompt, model="claude-sonnet-4-20250514")
+            response = client.analyze_sync(prompt, model=MODEL_AUXILIARY)
             result = _parse_json_response(response)
 
             if result and result.get("found_files"):

libs/openant-core/utilities/context_enhancer.py

@@ -27,6 +27,7 @@
 
 from .llm_client import AnthropicClient, TokenTracker, get_global_tracker, reset_global_tracker
 from .agentic_enhancer import RepositoryIndex, enhance_unit_with_agent, load_index_from_file
+from .model_config import MODEL_AUXILIARY
 from .rate_limiter import get_rate_limiter, is_rate_limit_error, is_retryable_error
 
 # Avoid circular import — import checkpoint at usage site
@@ -45,7 +46,7 @@ def _get_step_checkpoint():
 
 
 # Use Sonnet for context enhancement (cost-effective auxiliary task)
-CONTEXT_ENHANCEMENT_MODEL = "claude-sonnet-4-20250514"
+CONTEXT_ENHANCEMENT_MODEL = MODEL_AUXILIARY
 
 
 def _build_error_info(exc: Exception) -> dict:
@@ -568,7 +569,7 @@ def enhance_dataset_agentic(
         remaining = total - len(processed_ids)
         self._log("info", f"Enhancing {remaining} units with agentic analysis ({len(processed_ids)} already done)", units=remaining)
         self._log("info", "Mode: Iterative tool use (traces call paths)")
-        self._log("info", "Model: claude-sonnet-4-20250514")
+        self._log("info", f"Model: {CONTEXT_ENHANCEMENT_MODEL}")
         mode = "sequential" if workers <= 1 else f"parallel ({workers} workers)"
         self._log("info", f"Workers: {mode}")
         if checkpoint_dir: