Merge branch 'main' into kimi/issue-534

Co-authored-by: Kimi Agent <kimi@timmy.local>
Co-committed-by: Kimi Agent <kimi@timmy.local>

scripts/loop_guard.py

@@ -18,13 +18,19 @@ Exit codes:
 from __future__ import annotations
 
 import json
+import os
 import sys
 import time
+import urllib.request
 from pathlib import Path
 
 REPO_ROOT = Path(__file__).resolve().parent.parent
 QUEUE_FILE = REPO_ROOT / ".loop" / "queue.json"
 IDLE_STATE_FILE = REPO_ROOT / ".loop" / "idle_state.json"
+TOKEN_FILE = Path.home() / ".hermes" / "gitea_token"
+
+GITEA_API = os.environ.get("GITEA_API", "http://localhost:3000/api/v1")
+REPO_SLUG = os.environ.get("REPO_SLUG", "rockachopa/Timmy-time-dashboard")
 
 # Backoff sequence: 60s, 120s, 240s, 600s max
 BACKOFF_BASE = 60
@@ -32,19 +38,81 @@ BACKOFF_MAX = 600
 BACKOFF_MULTIPLIER = 2
 
 
+def _get_token() -> str:
+    """Read Gitea token from env or file."""
+    token = os.environ.get("GITEA_TOKEN", "").strip()
+    if not token and TOKEN_FILE.exists():
+        token = TOKEN_FILE.read_text().strip()
+    return token
+
+
+def _fetch_open_issue_numbers() -> set[int] | None:
+    """Fetch open issue numbers from Gitea. Returns None on failure."""
+    token = _get_token()
+    if not token:
+        return None
+    try:
+        numbers: set[int] = set()
+        page = 1
+        while True:
+            url = (
+                f"{GITEA_API}/repos/{REPO_SLUG}/issues"
+                f"?state=open&type=issues&limit=50&page={page}"
+            )
+            req = urllib.request.Request(url, headers={
+                "Authorization": f"token {token}",
+                "Accept": "application/json",
+            })
+            with urllib.request.urlopen(req, timeout=10) as resp:
+                data = json.loads(resp.read())
+            if not data:
+                break
+            for issue in data:
+                numbers.add(issue["number"])
+            if len(data) < 50:
+                break
+            page += 1
+        return numbers
+    except Exception:
+        return None
+
+
 def load_queue() -> list[dict]:
-    """Load queue.json and return ready items."""
+    """Load queue.json and return ready items, filtering out closed issues."""
     if not QUEUE_FILE.exists():
         return []
     try:
         data = json.loads(QUEUE_FILE.read_text())
-        if isinstance(data, list):
-            return [item for item in data if item.get("ready")]
-        return []
+        if not isinstance(data, list):
+            return []
+        ready = [item for item in data if item.get("ready")]
+        if not ready:
+            return []
+
+        # Filter out issues that are no longer open (auto-hygiene)
+        open_numbers = _fetch_open_issue_numbers()
+        if open_numbers is not None:
+            before = len(ready)
+            ready = [item for item in ready if item.get("issue") in open_numbers]
+            removed = before - len(ready)
+            if removed > 0:
+                print(f"[loop-guard] Filtered {removed} closed issue(s) from queue")
+                # Persist the cleaned queue so stale entries don't recur
+                _save_cleaned_queue(data, open_numbers)
+        return ready
     except (json.JSONDecodeError, OSError):
         return []
 
 
+def _save_cleaned_queue(full_queue: list[dict], open_numbers: set[int]) -> None:
+    """Rewrite queue.json without closed issues."""
+    cleaned = [item for item in full_queue if item.get("issue") in open_numbers]
+    try:
+        QUEUE_FILE.write_text(json.dumps(cleaned, indent=2) + "\n")
+    except OSError:
+        pass
+
+
 def load_idle_state() -> dict:
     """Load persistent idle state."""
     if not IDLE_STATE_FILE.exists():

src/config.py

@@ -254,6 +254,7 @@ class Settings(BaseSettings):
     # When enabled, the agent starts an internal thought loop on server start.
     thinking_enabled: bool = True
     thinking_interval_seconds: int = 300  # 5 minutes between thoughts
+    thinking_timeout_seconds: int = 120  # max wall-clock time per thinking cycle
     thinking_distill_every: int = 10  # distill facts from thoughts every Nth thought
     thinking_issue_every: int = 20  # file Gitea issues from thoughts every Nth thought
     thinking_memory_check_every: int = 50  # check memory status every Nth thought

src/dashboard/app.py

@@ -155,7 +155,17 @@ async def _thinking_scheduler() -> None:
     while True:
         try:
             if settings.thinking_enabled:
-                await thinking_engine.think_once()
+                await asyncio.wait_for(
+                    thinking_engine.think_once(),
+                    timeout=settings.thinking_timeout_seconds,
+                )
+        except TimeoutError:
+            logger.warning(
+                "Thinking cycle timed out after %ds — Ollama may be unresponsive",
+                settings.thinking_timeout_seconds,
+            )
+        except asyncio.CancelledError:
+            raise
         except Exception as exc:
             logger.error("Thinking scheduler error: %s", exc)
 
@@ -175,7 +185,10 @@ async def _loop_qa_scheduler() -> None:
     while True:
         try:
             if settings.loop_qa_enabled:
-                result = await loop_qa_orchestrator.run_next_test()
+                result = await asyncio.wait_for(
+                    loop_qa_orchestrator.run_next_test(),
+                    timeout=settings.thinking_timeout_seconds,
+                )
                 if result:
                     status = "PASS" if result["success"] else "FAIL"
                     logger.info(
@@ -184,6 +197,13 @@ async def _loop_qa_scheduler() -> None:
                         status,
                         result.get("details", "")[:80],
                     )
+        except TimeoutError:
+            logger.warning(
+                "Loop QA test timed out after %ds",
+                settings.thinking_timeout_seconds,
+            )
+        except asyncio.CancelledError:
+            raise
         except Exception as exc:
             logger.error("Loop QA scheduler error: %s", exc)
 
@@ -329,33 +349,35 @@ async def _discord_token_watcher() -> None:
                 logger.warning("Discord auto-start failed: %s", exc)
 
 
-@asynccontextmanager
-async def lifespan(app: FastAPI):
-    """Application lifespan manager with non-blocking startup."""
-
-    # Validate security config (no-op in test mode)
+def _startup_init() -> None:
+    """Validate config and enable event persistence."""
     from config import validate_startup
 
     validate_startup()
 
-    # Enable event persistence (unified EventBus + swarm event_log)
     from infrastructure.events.bus import init_event_bus_persistence
 
     init_event_bus_persistence()
 
-    # Create all background tasks without waiting for them
-    briefing_task = asyncio.create_task(_briefing_scheduler())
-    thinking_task = asyncio.create_task(_thinking_scheduler())
-    loop_qa_task = asyncio.create_task(_loop_qa_scheduler())
-    presence_task = asyncio.create_task(_presence_watcher())
-
-    # Initialize Spark Intelligence engine
     from spark.engine import get_spark_engine
 
     if get_spark_engine().enabled:
         logger.info("Spark Intelligence active — event capture enabled")
 
-    # Auto-prune old vector store memories on startup
+
+def _startup_background_tasks() -> list[asyncio.Task]:
+    """Spawn all recurring background tasks (non-blocking)."""
+    return [
+        asyncio.create_task(_briefing_scheduler()),
+        asyncio.create_task(_thinking_scheduler()),
+        asyncio.create_task(_loop_qa_scheduler()),
+        asyncio.create_task(_presence_watcher()),
+        asyncio.create_task(_start_chat_integrations_background()),
+    ]
+
+
+def _startup_pruning() -> None:
+    """Auto-prune old memories, thoughts, and events on startup."""
     if settings.memory_prune_days > 0:
         try:
             from timmy.memory_system import prune_memories
@@ -373,7 +395,6 @@ async def lifespan(app: FastAPI):
         except Exception as exc:
             logger.debug("Memory auto-prune skipped: %s", exc)
 
-    # Auto-prune old thoughts on startup
     if settings.thoughts_prune_days > 0:
         try:
             from timmy.thinking import thinking_engine
@@ -391,7 +412,6 @@ async def lifespan(app: FastAPI):
         except Exception as exc:
             logger.debug("Thought auto-prune skipped: %s", exc)
 
-    # Auto-prune old system events on startup
     if settings.events_prune_days > 0:
         try:
             from swarm.event_log import prune_old_events
@@ -409,7 +429,6 @@ async def lifespan(app: FastAPI):
         except Exception as exc:
             logger.debug("Event auto-prune skipped: %s", exc)
 
-    # Warn if memory vault exceeds size limit
     if settings.memory_vault_max_mb > 0:
         try:
             vault_path = Path(settings.repo_root) / "memory" / "notes"
@@ -425,6 +444,42 @@ async def lifespan(app: FastAPI):
         except Exception as exc:
             logger.debug("Vault size check skipped: %s", exc)
 
+
+async def _shutdown_cleanup(
+    bg_tasks: list[asyncio.Task],
+    workshop_heartbeat,
+) -> None:
+    """Stop chat bots, MCP sessions, heartbeat, and cancel background tasks."""
+    from integrations.chat_bridge.vendors.discord import discord_bot
+    from integrations.telegram_bot.bot import telegram_bot
+
+    await discord_bot.stop()
+    await telegram_bot.stop()
+
+    try:
+        from timmy.mcp_tools import close_mcp_sessions
+
+        await close_mcp_sessions()
+    except Exception as exc:
+        logger.debug("MCP shutdown: %s", exc)
+
+    await workshop_heartbeat.stop()
+
+    for task in bg_tasks:
+        task.cancel()
+        try:
+            await task
+        except asyncio.CancelledError:
+            pass
+
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    """Application lifespan manager with non-blocking startup."""
+    _startup_init()
+    bg_tasks = _startup_background_tasks()
+    _startup_pruning()
+
     # Start Workshop presence heartbeat with WS relay
     from dashboard.routes.world import broadcast_world_state
     from timmy.workshop_state import WorkshopHeartbeat
@@ -432,10 +487,7 @@ async def lifespan(app: FastAPI):
     workshop_heartbeat = WorkshopHeartbeat(on_change=broadcast_world_state)
     await workshop_heartbeat.start()
 
-    # Start chat integrations in background
-    chat_task = asyncio.create_task(_start_chat_integrations_background())
-
-    # Register session logger with error capture (breaks infrastructure → timmy circular dep)
+    # Register session logger with error capture
     try:
         from infrastructure.error_capture import register_error_recorder
         from timmy.session_logger import get_session_logger
@@ -448,30 +500,7 @@ async def lifespan(app: FastAPI):
 
     yield
 
-    # Cleanup on shutdown
-    from integrations.chat_bridge.vendors.discord import discord_bot
-    from integrations.telegram_bot.bot import telegram_bot
-
-    await discord_bot.stop()
-    await telegram_bot.stop()
-
-    # Close MCP tool server sessions
-    try:
-        from timmy.mcp_tools import close_mcp_sessions
-
-        await close_mcp_sessions()
-    except Exception as exc:
-        logger.debug("MCP shutdown: %s", exc)
-
-    await workshop_heartbeat.stop()
-
-    for task in [briefing_task, thinking_task, chat_task, loop_qa_task, presence_task]:
-        if task:
-            task.cancel()
-            try:
-                await task
-            except asyncio.CancelledError:
-                pass
+    await _shutdown_cleanup(bg_tasks, workshop_heartbeat)
 
 
 app = FastAPI(

src/infrastructure/error_capture.py

@@ -100,36 +100,14 @@ def _get_git_context() -> dict:
         return {"branch": "unknown", "commit": "unknown"}
 
 
-def capture_error(
-    exc: Exception,
-    source: str = "unknown",
-    context: dict | None = None,
-) -> str | None:
-    """Capture an error and optionally create a bug report.
-
-    Args:
-        exc: The exception to capture
-        source: Module/component where the error occurred
-        context: Optional dict of extra context (request path, etc.)
+def _extract_traceback_info(exc: Exception) -> tuple[str, str, int]:
+    """Extract formatted traceback, affected file, and line number.
 
     Returns:
-        Task ID of the created bug report, or None if deduplicated/disabled
+        Tuple of (traceback_string, affected_file, affected_line).
     """
-    from config import settings
-
-    if not settings.error_feedback_enabled:
-        return None
-
-    error_hash = _stack_hash(exc)
-
-    if _is_duplicate(error_hash):
-        logger.debug("Duplicate error suppressed: %s (hash=%s)", exc, error_hash)
-        return None
-
-    # Format the stack trace
     tb_str = "".join(traceback.format_exception(type(exc), exc, exc.__traceback__))
 
-    # Extract file/line from traceback
     tb_obj = exc.__traceback__
     affected_file = "unknown"
     affected_line = 0
@@ -139,9 +117,18 @@ def capture_error(
         affected_file = tb_obj.tb_frame.f_code.co_filename
         affected_line = tb_obj.tb_lineno
 
-    git_ctx = _get_git_context()
+    return tb_str, affected_file, affected_line
 
-    # 1. Log to event_log
+
+def _log_error_event(
+    exc: Exception,
+    source: str,
+    error_hash: str,
+    affected_file: str,
+    affected_line: int,
+    git_ctx: dict,
+) -> None:
+    """Log the captured error to the event log."""
     try:
         from swarm.event_log import EventType, log_event
 
@@ -161,8 +148,18 @@ def capture_error(
     except Exception as log_exc:
         logger.debug("Failed to log error event: %s", log_exc)
 
-    # 2. Create bug report task
-    task_id = None
+
+def _create_bug_report(
+    exc: Exception,
+    source: str,
+    context: dict | None,
+    error_hash: str,
+    tb_str: str,
+    affected_file: str,
+    affected_line: int,
+    git_ctx: dict,
+) -> str | None:
+    """Create a bug report task and return the task ID (or None on failure)."""
     try:
         from swarm.task_queue.models import create_task
 
@@ -195,7 +192,6 @@ def capture_error(
         )
         task_id = task.id
 
-        # Log the creation event
         try:
             from swarm.event_log import EventType, log_event
 
@@ -210,12 +206,16 @@ def capture_error(
             )
         except Exception as exc:
             logger.warning("Bug report screenshot error: %s", exc)
-            pass
+
+        return task_id
 
     except Exception as task_exc:
         logger.debug("Failed to create bug report task: %s", task_exc)
+        return None
 
-    # 3. Send notification
+
+def _notify_bug_report(exc: Exception, source: str) -> None:
+    """Send a push notification about the captured error."""
     try:
         from infrastructure.notifications.push import notifier
 
@@ -224,11 +224,12 @@ def capture_error(
             message=f"{type(exc).__name__} in {source}: {str(exc)[:80]}",
             category="system",
         )
-    except Exception as exc:
-        logger.warning("Bug report notification error: %s", exc)
-        pass
+    except Exception as notify_exc:
+        logger.warning("Bug report notification error: %s", notify_exc)
 
-    # 4. Record in session logger (via registered callback)
+
+def _record_to_session(exc: Exception, source: str) -> None:
+    """Record the error via the registered session callback."""
     if _error_recorder is not None:
         try:
             _error_recorder(
@@ -238,4 +239,50 @@ def capture_error(
         except Exception as log_exc:
             logger.warning("Bug report session logging error: %s", log_exc)
 
+
+def capture_error(
+    exc: Exception,
+    source: str = "unknown",
+    context: dict | None = None,
+) -> str | None:
+    """Capture an error and optionally create a bug report.
+
+    Args:
+        exc: The exception to capture
+        source: Module/component where the error occurred
+        context: Optional dict of extra context (request path, etc.)
+
+    Returns:
+        Task ID of the created bug report, or None if deduplicated/disabled
+    """
+    from config import settings
+
+    if not settings.error_feedback_enabled:
+        return None
+
+    error_hash = _stack_hash(exc)
+
+    if _is_duplicate(error_hash):
+        logger.debug("Duplicate error suppressed: %s (hash=%s)", exc, error_hash)
+        return None
+
+    tb_str, affected_file, affected_line = _extract_traceback_info(exc)
+    git_ctx = _get_git_context()
+
+    _log_error_event(exc, source, error_hash, affected_file, affected_line, git_ctx)
+
+    task_id = _create_bug_report(
+        exc,
+        source,
+        context,
+        error_hash,
+        tb_str,
+        affected_file,
+        affected_line,
+        git_ctx,
+    )
+
+    _notify_bug_report(exc, source)
+    _record_to_session(exc, source)
+
     return task_id

src/timmy/agent.py

@@ -197,6 +197,90 @@ def _resolve_backend(requested: str | None) -> str:
     return "ollama"
 
 
+def _build_tools_list(use_tools: bool, skip_mcp: bool, model_name: str) -> list:
+    """Assemble the tools list based on model capability and MCP flags.
+
+    Returns a list of Toolkit / MCPTools objects, or an empty list.
+    """
+    if not use_tools:
+        logger.info("Tools disabled for model %s (too small for reliable tool calling)", model_name)
+        return []
+
+    tools_list: list = [create_full_toolkit()]
+
+    # Add MCP tool servers (lazy-connected on first arun()).
+    # Skipped when skip_mcp=True — MCP's stdio transport uses anyio cancel
+    # scopes that conflict with asyncio background task cancellation (#72).
+    if not skip_mcp:
+        try:
+            from timmy.mcp_tools import create_filesystem_mcp_tools, create_gitea_mcp_tools
+
+            gitea_mcp = create_gitea_mcp_tools()
+            if gitea_mcp:
+                tools_list.append(gitea_mcp)
+
+            fs_mcp = create_filesystem_mcp_tools()
+            if fs_mcp:
+                tools_list.append(fs_mcp)
+        except Exception as exc:
+            logger.debug("MCP tools unavailable: %s", exc)
+
+    return tools_list
+
+
+def _build_prompt(use_tools: bool, session_id: str) -> str:
+    """Build the full system prompt with optional memory context."""
+    base_prompt = get_system_prompt(tools_enabled=use_tools, session_id=session_id)
+
+    try:
+        from timmy.memory_system import memory_system
+
+        memory_context = memory_system.get_system_context()
+        if memory_context:
+            # Smaller budget for small models — expanded prompt uses more tokens
+            max_context = 2000 if not use_tools else 8000
+            if len(memory_context) > max_context:
+                memory_context = memory_context[:max_context] + "\n... [truncated]"
+            return (
+                f"{base_prompt}\n\n"
+                f"## GROUNDED CONTEXT (verified sources — cite when using)\n\n"
+                f"{memory_context}"
+            )
+    except Exception as exc:
+        logger.warning("Failed to load memory context: %s", exc)
+
+    return base_prompt
+
+
+def _create_ollama_agent(
+    *,
+    db_file: str,
+    model_name: str,
+    tools_list: list,
+    full_prompt: str,
+    use_tools: bool,
+) -> Agent:
+    """Construct the Agno Agent with Ollama backend and warm up the model."""
+    model_kwargs = {}
+    if settings.ollama_num_ctx > 0:
+        model_kwargs["options"] = {"num_ctx": settings.ollama_num_ctx}
+
+    agent = Agent(
+        name="Agent",
+        model=Ollama(id=model_name, host=settings.ollama_url, timeout=300, **model_kwargs),
+        db=SqliteDb(db_file=db_file),
+        description=full_prompt,
+        add_history_to_context=True,
+        num_history_runs=20,
+        markdown=False,
+        tools=tools_list if tools_list else None,
+        tool_call_limit=settings.max_agent_steps if use_tools else None,
+        telemetry=settings.telemetry_enabled,
+    )
+    _warmup_model(model_name)
+    return agent
+
+
 def create_timmy(
     db_file: str = "timmy.db",
     backend: str | None = None,
@@ -238,16 +322,12 @@ def create_timmy(
         return TimmyAirLLMAgent(model_size=size)
 
     # Default: Ollama via Agno.
-    # Resolve model with automatic pulling and fallback
     model_name, is_fallback = _resolve_model_with_fallback(
         requested_model=None,
         require_vision=False,
         auto_pull=True,
     )
 
-    # If Ollama is completely unreachable, fail loudly.
-    # Sovereignty: never silently send data to a cloud API.
-    # Use --backend claude explicitly if you want cloud inference.
     if not _check_model_available(model_name):
         logger.error(
             "Ollama unreachable and no local models available. "
@@ -258,76 +338,16 @@ def create_timmy(
         logger.info("Using fallback model %s (requested was unavailable)", model_name)
 
     use_tools = _model_supports_tools(model_name)
+    tools_list = _build_tools_list(use_tools, skip_mcp, model_name)
+    full_prompt = _build_prompt(use_tools, session_id)
 
-    # Conditionally include tools — small models get none
-    toolkit = create_full_toolkit() if use_tools else None
-    if not use_tools:
-        logger.info("Tools disabled for model %s (too small for reliable tool calling)", model_name)
-
-    # Build the tools list — Agno accepts a list of Toolkit / MCPTools
-    tools_list: list = []
-    if toolkit:
-        tools_list.append(toolkit)
-
-    # Add MCP tool servers (lazy-connected on first arun()).
-    # Skipped when skip_mcp=True — MCP's stdio transport uses anyio cancel
-    # scopes that conflict with asyncio background task cancellation (#72).
-    if use_tools and not skip_mcp:
-        try:
-            from timmy.mcp_tools import create_filesystem_mcp_tools, create_gitea_mcp_tools
-
-            gitea_mcp = create_gitea_mcp_tools()
-            if gitea_mcp:
-                tools_list.append(gitea_mcp)
-
-            fs_mcp = create_filesystem_mcp_tools()
-            if fs_mcp:
-                tools_list.append(fs_mcp)
-        except Exception as exc:
-            logger.debug("MCP tools unavailable: %s", exc)
-
-    # Select prompt tier based on tool capability
-    base_prompt = get_system_prompt(tools_enabled=use_tools, session_id=session_id)
-
-    # Try to load memory context
-    try:
-        from timmy.memory_system import memory_system
-
-        memory_context = memory_system.get_system_context()
-        if memory_context:
-            # Truncate if too long — smaller budget for small models
-            # since the expanded prompt (roster, guardrails) uses more tokens
-            max_context = 2000 if not use_tools else 8000
-            if len(memory_context) > max_context:
-                memory_context = memory_context[:max_context] + "\n... [truncated]"
-            full_prompt = (
-                f"{base_prompt}\n\n"
-                f"## GROUNDED CONTEXT (verified sources — cite when using)\n\n"
-                f"{memory_context}"
-            )
-        else:
-            full_prompt = base_prompt
-    except Exception as exc:
-        logger.warning("Failed to load memory context: %s", exc)
-        full_prompt = base_prompt
-
-    model_kwargs = {}
-    if settings.ollama_num_ctx > 0:
-        model_kwargs["options"] = {"num_ctx": settings.ollama_num_ctx}
-    agent = Agent(
-        name="Agent",
-        model=Ollama(id=model_name, host=settings.ollama_url, timeout=300, **model_kwargs),
-        db=SqliteDb(db_file=db_file),
-        description=full_prompt,
-        add_history_to_context=True,
-        num_history_runs=20,
-        markdown=False,
-        tools=tools_list if tools_list else None,
-        tool_call_limit=settings.max_agent_steps if use_tools else None,
-        telemetry=settings.telemetry_enabled,
+    return _create_ollama_agent(
+        db_file=db_file,
+        model_name=model_name,
+        tools_list=tools_list,
+        full_prompt=full_prompt,
+        use_tools=use_tools,
     )
-    _warmup_model(model_name)
-    return agent
 
 
 class TimmyWithMemory:

src/timmy/thinking.py

@@ -232,6 +232,90 @@ class ThinkingEngine:
             return False  # Disabled — never idle
         return datetime.now(UTC) - self._last_input_time > timedelta(minutes=timeout)
 
+    def _build_thinking_context(self) -> tuple[str, str, list["Thought"]]:
+        """Assemble the context needed for a thinking cycle.
+
+        Returns:
+            (memory_context, system_context, recent_thoughts)
+        """
+        memory_context = self._load_memory_context()
+        system_context = self._gather_system_snapshot()
+        recent_thoughts = self.get_recent_thoughts(limit=5)
+        return memory_context, system_context, recent_thoughts
+
+    async def _generate_novel_thought(
+        self,
+        prompt: str | None,
+        memory_context: str,
+        system_context: str,
+        recent_thoughts: list["Thought"],
+    ) -> tuple[str | None, str]:
+        """Run the dedup-retry loop to produce a novel thought.
+
+        Returns:
+            (content, seed_type) — content is None if no novel thought produced.
+        """
+        seed_type: str = "freeform"
+
+        for attempt in range(self._MAX_DEDUP_RETRIES + 1):
+            if prompt:
+                seed_type = "prompted"
+                seed_context = f"Journal prompt: {prompt}"
+            else:
+                seed_type, seed_context = self._gather_seed()
+
+            continuity = self._build_continuity_context()
+
+            full_prompt = _THINKING_PROMPT.format(
+                memory_context=memory_context,
+                system_context=system_context,
+                seed_context=seed_context,
+                continuity_context=continuity,
+            )
+
+            try:
+                raw = await self._call_agent(full_prompt)
+            except Exception as exc:
+                logger.warning("Thinking cycle failed (Ollama likely down): %s", exc)
+                return None, seed_type
+
+            if not raw or not raw.strip():
+                logger.debug("Thinking cycle produced empty response, skipping")
+                return None, seed_type
+
+            content = raw.strip()
+
+            # Dedup: reject thoughts too similar to recent ones
+            if not self._is_too_similar(content, recent_thoughts):
+                return content, seed_type  # Good — novel thought
+
+            if attempt < self._MAX_DEDUP_RETRIES:
+                logger.info(
+                    "Thought too similar to recent (attempt %d/%d), retrying with new seed",
+                    attempt + 1,
+                    self._MAX_DEDUP_RETRIES + 1,
+                )
+            else:
+                logger.warning(
+                    "Thought still repetitive after %d retries, discarding",
+                    self._MAX_DEDUP_RETRIES + 1,
+                )
+                return None, seed_type
+
+        return None, seed_type
+
+    async def _process_thinking_result(self, thought: "Thought") -> None:
+        """Run all post-hooks after a thought is stored."""
+        self._maybe_check_memory()
+        await self._maybe_distill()
+        await self._maybe_file_issues()
+        await self._check_workspace()
+        self._maybe_check_memory_status()
+        self._update_memory(thought)
+        self._log_event(thought)
+        self._write_journal(thought)
+        await self._broadcast(thought)
+
     async def think_once(self, prompt: str | None = None) -> Thought | None:
         """Execute one thinking cycle.
 
@@ -257,91 +341,21 @@ class ThinkingEngine:
             )
             return None
 
-        memory_context = self._load_memory_context()
-        system_context = self._gather_system_snapshot()
-        recent_thoughts = self.get_recent_thoughts(limit=5)
-
-        content: str | None = None
-        seed_type: str = "freeform"
-
-        for attempt in range(self._MAX_DEDUP_RETRIES + 1):
-            if prompt:
-                seed_type = "prompted"
-                seed_context = f"Journal prompt: {prompt}"
-            else:
-                seed_type, seed_context = self._gather_seed()
-
-            continuity = self._build_continuity_context()
-
-            full_prompt = _THINKING_PROMPT.format(
-                memory_context=memory_context,
-                system_context=system_context,
-                seed_context=seed_context,
-                continuity_context=continuity,
-            )
-
-            try:
-                raw = await self._call_agent(full_prompt)
-            except Exception as exc:
-                logger.warning("Thinking cycle failed (Ollama likely down): %s", exc)
-                return None
-
-            if not raw or not raw.strip():
-                logger.debug("Thinking cycle produced empty response, skipping")
-                return None
-
-            content = raw.strip()
-
-            # Dedup: reject thoughts too similar to recent ones
-            if not self._is_too_similar(content, recent_thoughts):
-                break  # Good — novel thought
-
-            if attempt < self._MAX_DEDUP_RETRIES:
-                logger.info(
-                    "Thought too similar to recent (attempt %d/%d), retrying with new seed",
-                    attempt + 1,
-                    self._MAX_DEDUP_RETRIES + 1,
-                )
-                content = None  # Will retry
-            else:
-                logger.warning(
-                    "Thought still repetitive after %d retries, discarding",
-                    self._MAX_DEDUP_RETRIES + 1,
-                )
-                return None
+        memory_context, system_context, recent_thoughts = self._build_thinking_context()
 
+        content, seed_type = await self._generate_novel_thought(
+            prompt,
+            memory_context,
+            system_context,
+            recent_thoughts,
+        )
         if not content:
             return None
 
         thought = self._store_thought(content, seed_type)
         self._last_thought_id = thought.id
 
-        # Post-hook: check memory status periodically
-        self._maybe_check_memory()
-
-        # Post-hook: distill facts from recent thoughts periodically
-        await self._maybe_distill()
-
-        # Post-hook: file Gitea issues for actionable observations
-        await self._maybe_file_issues()
-
-        # Post-hook: check workspace for new messages from Hermes
-        await self._check_workspace()
-
-        # Post-hook: proactive memory status audit
-        self._maybe_check_memory_status()
-
-        # Post-hook: update MEMORY.md with latest reflection
-        self._update_memory(thought)
-
-        # Log to swarm event system
-        self._log_event(thought)
-
-        # Append to daily journal file
-        self._write_journal(thought)
-
-        # Broadcast to WebSocket clients
-        await self._broadcast(thought)
+        await self._process_thinking_result(thought)
 
         logger.info(
             "Thought [%s] (%s): %s",
@@ -1110,21 +1124,37 @@ class ThinkingEngine:
             lines.append(f"- [{thought.seed_type}] {snippet}")
         return "\n".join(lines)
 
+    _thinking_agent = None  # cached agent — avoids per-call resource leaks (#525)
+
     async def _call_agent(self, prompt: str) -> str:
         """Call Timmy's agent to generate a thought.
 
-        Creates a lightweight agent with skip_mcp=True to avoid the cancel-scope
+        Reuses a cached agent with skip_mcp=True to avoid the cancel-scope
         errors that occur when MCP stdio transports are spawned inside asyncio
-        background tasks (#72).  The thinking engine doesn't need Gitea or
-        filesystem tools — it only needs the LLM.
+        background tasks (#72) and to prevent per-call resource leaks (httpx
+        clients, SQLite connections, model warmups) that caused the thinking
+        loop to die every ~10 min (#525).
+
+        Individual calls are capped at 120 s so a hung Ollama never blocks
+        the scheduler indefinitely.
 
         Strips ``<think>`` tags from reasoning models (qwen3, etc.) so that
         downstream parsers (fact distillation, issue filing) receive clean text.
         """
-        from timmy.agent import create_timmy
+        import asyncio
+
+        if self._thinking_agent is None:
+            from timmy.agent import create_timmy
+
+            self._thinking_agent = create_timmy(skip_mcp=True)
+
+        try:
+            async with asyncio.timeout(120):
+                run = await self._thinking_agent.arun(prompt, stream=False)
+        except TimeoutError:
+            logger.warning("Thinking LLM call timed out after 120 s")
+            return ""
 
-        agent = create_timmy(skip_mcp=True)
-        run = await agent.arun(prompt, stream=False)
         raw = run.content if hasattr(run, "content") else str(run)
         return _THINK_TAG_RE.sub("", raw) if raw else raw
 

src/timmy/tools_intro/__init__.py

@@ -127,54 +127,48 @@ def check_ollama_health() -> dict[str, Any]:
     return result
 
 
-def get_memory_status() -> dict[str, Any]:
-    """Get the status of Timmy's memory system.
-
-    Returns:
-        Dict with memory tier information
-    """
-    from config import settings
-
-    repo_root = Path(settings.repo_root)
-
-    # Check tier 1: Hot memory
+def _hot_memory_info(repo_root: Path) -> dict[str, Any]:
+    """Tier 1: Hot memory (MEMORY.md) status."""
     memory_md = repo_root / "MEMORY.md"
     tier1_exists = memory_md.exists()
     tier1_content = ""
     if tier1_exists:
-        tier1_content = memory_md.read_text()[:500]  # First 500 chars
+        tier1_content = memory_md.read_text()[:500]
 
-    # Check tier 2: Vault
-    vault_path = repo_root / "memory" / "self"
-    tier2_exists = vault_path.exists()
-    tier2_files = []
-    if tier2_exists:
-        tier2_files = [f.name for f in vault_path.iterdir() if f.is_file()]
-
-    tier1_info: dict[str, Any] = {
+    info: dict[str, Any] = {
         "exists": tier1_exists,
         "path": str(memory_md),
         "preview": " ".join(tier1_content[:200].split()) if tier1_content else None,
     }
     if tier1_exists:
         lines = memory_md.read_text().splitlines()
-        tier1_info["line_count"] = len(lines)
-        tier1_info["sections"] = [ln.lstrip("# ").strip() for ln in lines if ln.startswith("## ")]
+        info["line_count"] = len(lines)
+        info["sections"] = [ln.lstrip("# ").strip() for ln in lines if ln.startswith("## ")]
+    return info
+
+
+def _vault_info(repo_root: Path) -> dict[str, Any]:
+    """Tier 2: Vault (memory/ directory tree) status."""
+    vault_path = repo_root / "memory" / "self"
+    tier2_exists = vault_path.exists()
+    tier2_files = [f.name for f in vault_path.iterdir() if f.is_file()] if tier2_exists else []
 
-    # Vault — scan all subdirs under memory/
     vault_root = repo_root / "memory"
-    vault_info: dict[str, Any] = {
+    info: dict[str, Any] = {
         "exists": tier2_exists,
         "path": str(vault_path),
         "file_count": len(tier2_files),
         "files": tier2_files[:10],
     }
     if vault_root.exists():
-        vault_info["directories"] = [d.name for d in vault_root.iterdir() if d.is_dir()]
-        vault_info["total_markdown_files"] = sum(1 for _ in vault_root.rglob("*.md"))
+        info["directories"] = [d.name for d in vault_root.iterdir() if d.is_dir()]
+        info["total_markdown_files"] = sum(1 for _ in vault_root.rglob("*.md"))
+    return info
 
-    # Tier 3: Semantic memory row count
-    tier3_info: dict[str, Any] = {"available": False}
+
+def _semantic_memory_info(repo_root: Path) -> dict[str, Any]:
+    """Tier 3: Semantic memory (vector DB) status."""
+    info: dict[str, Any] = {"available": False}
     try:
         sem_db = repo_root / "data" / "memory.db"
         if sem_db.exists():
@@ -184,14 +178,16 @@ def get_memory_status() -> dict[str, Any]:
                 ).fetchone()
                 if row and row[0]:
                     count = conn.execute("SELECT COUNT(*) FROM chunks").fetchone()
-                    tier3_info["available"] = True
-                    tier3_info["vector_count"] = count[0] if count else 0
+                    info["available"] = True
+                    info["vector_count"] = count[0] if count else 0
     except Exception as exc:
         logger.debug("Memory status query failed: %s", exc)
-        pass
+    return info
 
-    # Self-coding journal stats
-    journal_info: dict[str, Any] = {"available": False}
+
+def _journal_info(repo_root: Path) -> dict[str, Any]:
+    """Self-coding journal statistics."""
+    info: dict[str, Any] = {"available": False}
     try:
         journal_db = repo_root / "data" / "self_coding.db"
         if journal_db.exists():
@@ -203,7 +199,7 @@ def get_memory_status() -> dict[str, Any]:
                 if rows:
                     counts = {r["outcome"]: r["cnt"] for r in rows}
                     total = sum(counts.values())
-                    journal_info = {
+                    info = {
                         "available": True,
                         "total_attempts": total,
                         "successes": counts.get("success", 0),
@@ -212,13 +208,24 @@ def get_memory_status() -> dict[str, Any]:
                     }
     except Exception as exc:
         logger.debug("Journal stats query failed: %s", exc)
-        pass
+    return info
+
+
+def get_memory_status() -> dict[str, Any]:
+    """Get the status of Timmy's memory system.
+
+    Returns:
+        Dict with memory tier information
+    """
+    from config import settings
+
+    repo_root = Path(settings.repo_root)
 
     return {
-        "tier1_hot_memory": tier1_info,
-        "tier2_vault": vault_info,
-        "tier3_semantic": tier3_info,
-        "self_coding_journal": journal_info,
+        "tier1_hot_memory": _hot_memory_info(repo_root),
+        "tier2_vault": _vault_info(repo_root),
+        "tier3_semantic": _semantic_memory_info(repo_root),
+        "self_coding_journal": _journal_info(repo_root),
     }
 
 

tests/infrastructure/test_error_capture.py

@@ -5,9 +5,14 @@ from datetime import UTC, datetime, timedelta
 from unittest.mock import patch
 
 from infrastructure.error_capture import (
+    _create_bug_report,
     _dedup_cache,
+    _extract_traceback_info,
     _get_git_context,
     _is_duplicate,
+    _log_error_event,
+    _notify_bug_report,
+    _record_to_session,
     _stack_hash,
     capture_error,
 )
@@ -193,3 +198,91 @@ class TestCaptureError:
 
     def teardown_method(self):
         _dedup_cache.clear()
+
+
+class TestExtractTracebackInfo:
+    """Test _extract_traceback_info helper."""
+
+    def test_returns_three_tuple(self):
+        try:
+            raise ValueError("extract test")
+        except ValueError as e:
+            tb_str, affected_file, affected_line = _extract_traceback_info(e)
+        assert "ValueError" in tb_str
+        assert "extract test" in tb_str
+        assert affected_file.endswith(".py")
+        assert affected_line > 0
+
+    def test_file_points_to_raise_site(self):
+        try:
+            _make_exception()
+        except ValueError as e:
+            _, affected_file, _ = _extract_traceback_info(e)
+        assert "test_error_capture" in affected_file
+
+
+class TestLogErrorEvent:
+    """Test _log_error_event helper."""
+
+    def test_does_not_crash_on_missing_deps(self):
+        try:
+            raise RuntimeError("log test")
+        except RuntimeError as e:
+            _log_error_event(e, "test", "abc123", "file.py", 42, {"branch": "main"})
+
+
+class TestCreateBugReport:
+    """Test _create_bug_report helper."""
+
+    def test_does_not_crash_on_missing_deps(self):
+        try:
+            raise RuntimeError("report test")
+        except RuntimeError as e:
+            result = _create_bug_report(
+                e, "test", None, "abc123", "traceback...", "file.py", 42, {}
+            )
+        # May return None if swarm deps unavailable — that's fine
+        assert result is None or isinstance(result, str)
+
+    def test_with_context(self):
+        try:
+            raise RuntimeError("ctx test")
+        except RuntimeError as e:
+            result = _create_bug_report(e, "test", {"path": "/api"}, "abc", "tb", "f.py", 1, {})
+        assert result is None or isinstance(result, str)
+
+
+class TestNotifyBugReport:
+    """Test _notify_bug_report helper."""
+
+    def test_does_not_crash(self):
+        try:
+            raise RuntimeError("notify test")
+        except RuntimeError as e:
+            _notify_bug_report(e, "test")
+
+
+class TestRecordToSession:
+    """Test _record_to_session helper."""
+
+    def test_does_not_crash_without_recorder(self):
+        try:
+            raise RuntimeError("session test")
+        except RuntimeError as e:
+            _record_to_session(e, "test")
+
+    def test_calls_registered_recorder(self):
+        from infrastructure.error_capture import register_error_recorder
+
+        calls = []
+        register_error_recorder(lambda **kwargs: calls.append(kwargs))
+        try:
+            try:
+                raise RuntimeError("callback test")
+            except RuntimeError as e:
+                _record_to_session(e, "test_source")
+            assert len(calls) == 1
+            assert "RuntimeError" in calls[0]["error"]
+            assert calls[0]["context"] == "test_source"
+        finally:
+            register_error_recorder(None)

tests/timmy/test_agent.py

@@ -444,6 +444,150 @@ def test_get_effective_ollama_model_walks_fallback_chain():
         assert result == "fb-2"
 
 
+# ── _build_tools_list ─────────────────────────────────────────────────────
+
+
+def test_build_tools_list_empty_when_tools_disabled():
+    """Small models get an empty tools list."""
+    from timmy.agent import _build_tools_list
+
+    result = _build_tools_list(use_tools=False, skip_mcp=False, model_name="llama3.2")
+    assert result == []
+
+
+def test_build_tools_list_includes_toolkit_when_enabled():
+    """Tool-capable models get the full toolkit."""
+    mock_toolkit = MagicMock()
+    with patch("timmy.agent.create_full_toolkit", return_value=mock_toolkit):
+        from timmy.agent import _build_tools_list
+
+        result = _build_tools_list(use_tools=True, skip_mcp=True, model_name="llama3.1")
+        assert mock_toolkit in result
+
+
+def test_build_tools_list_skips_mcp_when_flagged():
+    """skip_mcp=True must not call MCP factories."""
+    mock_toolkit = MagicMock()
+    with (
+        patch("timmy.agent.create_full_toolkit", return_value=mock_toolkit),
+        patch("timmy.mcp_tools.create_gitea_mcp_tools") as mock_gitea,
+        patch("timmy.mcp_tools.create_filesystem_mcp_tools") as mock_fs,
+    ):
+        from timmy.agent import _build_tools_list
+
+        _build_tools_list(use_tools=True, skip_mcp=True, model_name="llama3.1")
+        mock_gitea.assert_not_called()
+        mock_fs.assert_not_called()
+
+
+def test_build_tools_list_includes_mcp_when_not_skipped():
+    """skip_mcp=False should attempt MCP tool creation."""
+    mock_toolkit = MagicMock()
+    with (
+        patch("timmy.agent.create_full_toolkit", return_value=mock_toolkit),
+        patch("timmy.mcp_tools.create_gitea_mcp_tools", return_value=None) as mock_gitea,
+        patch("timmy.mcp_tools.create_filesystem_mcp_tools", return_value=None) as mock_fs,
+    ):
+        from timmy.agent import _build_tools_list
+
+        _build_tools_list(use_tools=True, skip_mcp=False, model_name="llama3.1")
+        mock_gitea.assert_called_once()
+        mock_fs.assert_called_once()
+
+
+# ── _build_prompt ─────────────────────────────────────────────────────────
+
+
+def test_build_prompt_includes_base_prompt():
+    """Prompt should always contain the base system prompt."""
+    from timmy.agent import _build_prompt
+
+    result = _build_prompt(use_tools=False, session_id="test")
+    assert "Timmy" in result
+
+
+def test_build_prompt_appends_memory_context():
+    """Memory context should be appended when available."""
+    mock_memory = MagicMock()
+    mock_memory.get_system_context.return_value = "User prefers dark mode."
+    with patch("timmy.memory_system.memory_system", mock_memory):
+        from timmy.agent import _build_prompt
+
+        result = _build_prompt(use_tools=True, session_id="test")
+        assert "GROUNDED CONTEXT" in result
+        assert "dark mode" in result
+
+
+def test_build_prompt_truncates_long_memory():
+    """Long memory context should be truncated."""
+    mock_memory = MagicMock()
+    mock_memory.get_system_context.return_value = "x" * 10000
+    with patch("timmy.memory_system.memory_system", mock_memory):
+        from timmy.agent import _build_prompt
+
+        result = _build_prompt(use_tools=False, session_id="test")
+        assert "[truncated]" in result
+
+
+def test_build_prompt_survives_memory_failure():
+    """Prompt should fall back to base when memory fails."""
+    mock_memory = MagicMock()
+    mock_memory.get_system_context.side_effect = RuntimeError("db locked")
+    with patch("timmy.memory_system.memory_system", mock_memory):
+        from timmy.agent import _build_prompt
+
+        result = _build_prompt(use_tools=True, session_id="test")
+        assert "Timmy" in result
+        # Memory context should NOT be appended (the db locked error was caught)
+        assert "db locked" not in result
+
+
+# ── _create_ollama_agent ──────────────────────────────────────────────────
+
+
+def test_create_ollama_agent_passes_correct_kwargs():
+    """_create_ollama_agent must pass the expected kwargs to Agent."""
+    with (
+        patch("timmy.agent.Agent") as MockAgent,
+        patch("timmy.agent.Ollama"),
+        patch("timmy.agent.SqliteDb"),
+        patch("timmy.agent._warmup_model", return_value=True),
+    ):
+        from timmy.agent import _create_ollama_agent
+
+        _create_ollama_agent(
+            db_file="test.db",
+            model_name="llama3.1",
+            tools_list=[MagicMock()],
+            full_prompt="test prompt",
+            use_tools=True,
+        )
+        kwargs = MockAgent.call_args.kwargs
+        assert kwargs["description"] == "test prompt"
+        assert kwargs["markdown"] is False
+
+
+def test_create_ollama_agent_none_tools_when_empty():
+    """Empty tools_list should pass tools=None to Agent."""
+    with (
+        patch("timmy.agent.Agent") as MockAgent,
+        patch("timmy.agent.Ollama"),
+        patch("timmy.agent.SqliteDb"),
+        patch("timmy.agent._warmup_model", return_value=True),
+    ):
+        from timmy.agent import _create_ollama_agent
+
+        _create_ollama_agent(
+            db_file="test.db",
+            model_name="llama3.2",
+            tools_list=[],
+            full_prompt="test prompt",
+            use_tools=False,
+        )
+        kwargs = MockAgent.call_args.kwargs
+        assert kwargs["tools"] is None
+
+
 def test_no_hardcoded_fallback_constants_in_agent():
     """agent.py must not define module-level DEFAULT_MODEL_FALLBACKS."""
     import timmy.agent as agent_mod