utim-cli 2.1.3__tar.gz → 2.2.0__tar.gz

{utim_cli-2.1.3 → utim_cli-2.2.0}/CHANGELOG.md

@@ -1,9 +1,9 @@
 # Changelog
 
-## [2.1.3] - 2026-08-12
+## [2.2.0] - 2026-08-13
 
 ### 🛠️ Fixes
-- Minor modifications and Bug fixes
+- Modified interface
 
 ## [2.1.2] - 2026-08-10
 

{utim_cli-2.1.3/utim_cli.egg-info → utim_cli-2.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: utim-cli
-Version: 2.1.3
+Version: 2.2.0
 Summary: UTIM – Universal Terminal Intelligence Manager. A powerful agentic AI coding assistant for your terminal.
 License: MIT
 Project-URL: Homepage, https://utim.dev

{utim_cli-2.1.3 → utim_cli-2.2.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "utim-cli"
-version = "2.1.3"
+version = "2.2.0"
 description = "UTIM – Universal Terminal Intelligence Manager. A powerful agentic AI coding assistant for your terminal."
 readme = "README.md"
 requires-python = ">=3.9"

{utim_cli-2.1.3 → utim_cli-2.2.0}/utim_cli/__init__.py

@@ -5,7 +5,7 @@ try:
     from utim_cli._version import VERSION as __version__
 except ImportError:
     # Fallback during first-install / editable-install edge cases
-    __version__ = "2.1.3"
+    __version__ = "2.2.0"
 
 import builtins, os, pathlib, urllib.parse, re
 

{utim_cli-2.1.3 → utim_cli-2.2.0}/utim_cli/_version.py

@@ -11,4 +11,4 @@
 # from here so every surface stays in sync automatically.
 # ─────────────────────────────────────────────────────────────────────────────
 
-VERSION = "2.1.3"
+VERSION = "2.2.0"

{utim_cli-2.1.3 → utim_cli-2.2.0}/utim_cli/brain.py

@@ -83,6 +83,12 @@ BRAIN_FINAL_LIMIT     = 6      # max memories injected into system prompt
 BRAIN_RELEVANCE_FLOOR = 0.30   # cosine similarity floor for injection
 QWEN_BRAIN_MODEL      = "qwen/qwen2.5-1.5b-instruct"   # fast reasoning model on server
 
+# Reuse the reflection model list so brain calls use the same fallback chain
+try:
+    from utim_cli.reflection import REFLECTION_MODELS
+except Exception:
+    REFLECTION_MODELS = [QWEN_BRAIN_MODEL, "openrouter/free"]
+
 # ── Paths ──────────────────────────────────────────────────────────────────────
 
 def _brain_root() -> Path:
@@ -177,66 +183,50 @@ def _cosine(v1: List[float], v2: List[float]) -> float:
 
 def _call_qwen_brain(prompt: str, system: str = "", max_tokens: int = 400) -> str:
     """
-    Call Qwen3-0.6B via the UTIM server or OpenRouter for fast brain reasoning.
+    Call a lightweight free model on OpenRouter for internal reasoning tasks.
     Returns the raw text response, or "" on failure.
     """
+    import re as _re
     try:
-        from utim_cli.config import config
-        from utim_cli.auth import SERVER_URL
-        import requests as _req
-
-        api_key = config.get("api_key")
-        llm_key = os.getenv("OPENROUTER_API_KEY") or os.getenv("UTIM_API_KEY") or api_key
-
-        messages = []
-        if system:
-            messages.append({"role": "system", "content": system})
-        messages.append({"role": "user", "content": prompt})
-
-        if api_key:
-            payload = {
-                "messages": messages,
-                "model_id": QWEN_BRAIN_MODEL,
-                "max_tokens": max_tokens,
-                "tools": None,
-                "session_id": None,
-                "is_reflection": True,
-            }
-            resp = _req.post(
-                f"{SERVER_URL}/completions",
-                json=payload,
-                headers={"X-API-Key": api_key, "Content-Type": "application/json"},
-                timeout=20,
-            )
-            if resp.status_code == 200:
-                final = ""
-                for line in resp.iter_lines():
-                    if not line:
-                        continue
-                    try:
-                        ev = json.loads(line.decode("utf-8"))
-                        if ev.get("type") == "content_delta":
-                            final += ev.get("text") or ""
-                        elif ev.get("type") == "done" and not final:
-                            final = ev.get("content") or ""
-                    except Exception:
-                        pass
-                return _strip_think_tags(final)
-
-        # Fallback: direct OpenRouter
-        if llm_key:
+        from utim_cli.config import config as _cfg
+        _key = _cfg.get("api_key") or os.getenv("OPENROUTER_API_KEY") or os.getenv("UTIM_API_KEY") or ""
+    except Exception:
+        _key = os.getenv("OPENROUTER_API_KEY") or os.getenv("UTIM_API_KEY") or ""
+
+    if not _key:
+        return ""
+
+    import requests as _req
+    _system = system or "You are a precise AI assistant. Be concise and accurate."
+
+    for model in REFLECTION_MODELS:
+        try:
             resp = _req.post(
                 "https://openrouter.ai/api/v1/chat/completions",
-                json={"model": QWEN_BRAIN_MODEL, "messages": messages, "max_tokens": max_tokens},
-                headers={"Authorization": f"Bearer {llm_key}"},
+                headers={
+                    "Authorization": f"Bearer {_key}",
+                    "Content-Type": "application/json",
+                    "HTTP-Referer": "https://utim.dev",
+                    "X-Title": "UTIM Brain",
+                },
+                json={
+                    "model": model,
+                    "messages": [
+                        {"role": "system", "content": _system},
+                        {"role": "user", "content": prompt},
+                    ],
+                    "max_tokens": max_tokens,
+                    "temperature": 0.1,
+                },
                 timeout=20,
             )
-            if resp.status_code == 200:
-                return _strip_think_tags(
-                    resp.json()["choices"][0]["message"]["content"]
-                )
-    except Exception:
-        pass
+            if resp.status_code != 200:
+                continue
+            raw = resp.json().get("choices", [{}])[0].get("message", {}).get("content", "")
+            raw = _re.sub(r"<think(?:ing)?>.*?</think(?:ing)?>", "", raw, flags=_re.DOTALL).strip()
+            return raw
+        except Exception:
+            continue
     return ""
 
 def _strip_think_tags(text: str) -> str: