chore: bump version to 1.1.0a3

* fix: improve web_fetch reliability — strip scripts, dynamic truncation, more tokens

- strip_html() now removes <script>, <style>, <template>, <noscript>
  element content instead of just their tags
- Truncation budget scales with context window (75% in chars, 50k floor)
  and takes from the beginning only instead of head+tail splice
- max_tokens bumped from 2000 to 8192 so thinking models don't starve
  the visible extraction answer
- reasoning_effort="low" on summarization call to avoid wasting tokens
- Empty responses and empty extractions now report as tool errors

* refactor: extract _utility_completion to fix reasoning_effort duplication

Callers previously had to pass reasoning_effort both as a direct keyword
(for commercial providers) and via _provider_extra_params (for local
model servers).  This duplication was easy to get wrong — web_fetch was
already missing the direct keyword.

_utility_completion threads it through both paths from a single call,
used by title generation, compaction, and web_fetch extraction.

* fix: disable thinking when max_tokens too small, cap extraction at 500k

_reasoning_params now returns empty dict when max_tokens can't fit a
thinking budget (e.g. title gen with max_tokens=200).  Previously
produced budget_tokens >= max_tokens which is an API error on
manual-thinking Anthropic models.

Also caps web_fetch content truncation at 500k chars — the dynamic
context-window calc was producing 3M chars on 1M-context models.

* fix: clamp utility max_tokens to model output limit, add strip_html tests

_utility_completion now clamps max_tokens to the model's advertised
max_output_tokens so small/local models don't reject 8192-token
requests.

Adds 8 tests for invisible element stripping (script, style, template,
noscript) including multiline, case-insensitive, and attribute cases.

* fix: mock get_capabilities in title retry tests for _utility_completion

_utility_completion calls _get_capabilities to clamp max_tokens.  The
existing title tests mocked _provider as a bare MagicMock, so
caps.max_output_tokens was a truthy MagicMock instead of an int.  Set
get_capabilities to return a real ModelCapabilities instance.

.github/workflows/ci.yml

@@ -74,6 +74,53 @@ jobs:
         env:
           TURNSTONE_TEST_PG_URL: postgresql+psycopg://postgres:postgres@localhost:5432/turnstone_test
 
+  wheel-completeness:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6
+      - uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405 # v6
+        with:
+          python-version: "3.14"
+      - run: pip install build
+      - run: python -m build --wheel
+      - name: Check all data files are in wheel
+        run: |
+          SOURCE=$(find turnstone -type f \
+            ! -name '*.py' ! -name '*.pyc' ! -path '*__pycache__*' \
+            | sort)
+          WHEEL=$(python -m zipfile -l dist/*.whl \
+            | awk '{print $1}' \
+            | grep -v '\.py$' | grep -v '\.dist-info' | grep -v '\.pyc' | grep -v '^File$' \
+            | sort)
+
+          # Files intentionally excluded from the wheel (one per line)
+          ALLOW="
+          turnstone/core/storage/migrations/script.py.mako
+          "
+
+          MISSING=$(comm -23 <(echo "$SOURCE") <(echo "$WHEEL") \
+            | grep -vFxf <(echo "$ALLOW" | sed '/^[[:space:]]*$/d; s/^[[:space:]]*//' ) || true)
+
+          if [ -n "$MISSING" ]; then
+            echo "::error::Data files in source tree but missing from wheel:"
+            echo "$MISSING"
+            echo ""
+            echo "Add them to [tool.hatch.build.targets.wheel] in pyproject.toml"
+            echo "or to the ALLOW list in this job if intentionally excluded."
+            exit 1
+          fi
+          echo "All source data files present in wheel"
+      - name: Smoke-test entry points from installed wheel
+        run: |
+          python -m venv /tmp/smoke
+          /tmp/smoke/bin/pip install dist/*.whl
+          /tmp/smoke/bin/turnstone --help
+          /tmp/smoke/bin/turnstone-server --help
+          /tmp/smoke/bin/turnstone-console --help
+          /tmp/smoke/bin/turnstone-admin --help
+          /tmp/smoke/bin/turnstone-channel --help
+          /tmp/smoke/bin/turnstone-bootstrap --help
+
   lock-check:
     runs-on: ubuntu-latest
     steps:

.github/workflows/docker-publish.yml

@@ -5,6 +5,10 @@ on:
     workflows: ["CI"]
     types: [completed]
 
+concurrency:
+  group: docker-${{ github.event.workflow_run.head_sha }}
+  cancel-in-progress: true
+
 permissions:
   contents: read
   packages: write
@@ -37,7 +41,7 @@ jobs:
 
       - name: Log in to GHCR
         if: steps.tag.outputs.skip == 'false'
-        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3
+        uses: docker/login-action@4907a6ddec9925e35a0a9e82d7399ccc52663121 # v4
         with:
           registry: ${{ env.REGISTRY }}
           username: ${{ github.actor }}
@@ -61,12 +65,12 @@ jobs:
           fi
           echo "tags=${TAGS}" >> "$GITHUB_OUTPUT"
 
-      - uses: docker/setup-buildx-action@b5ca514318bd6ebac0fb2aedd5d36ec1b5c232a2 # v3
+      - uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4
         if: steps.tag.outputs.skip == 'false'
 
       - name: Build and push
         if: steps.tag.outputs.skip == 'false'
-        uses: docker/build-push-action@14487ce63c7a62a4a324b0bfb37086795e31c6c1 # v6
+        uses: docker/build-push-action@d08e5c354a6adb9ed34480a06d141179aa583294 # v7
         with:
           context: .
           push: true

.github/workflows/publish.yml

@@ -5,6 +5,10 @@ on:
     workflows: ["CI"]
     types: [completed]
 
+concurrency:
+  group: publish-${{ github.event.workflow_run.head_sha }}
+  cancel-in-progress: true
+
 permissions:
   contents: write
   id-token: write

compose.yaml

@@ -60,9 +60,7 @@ services:
   # turnstone-server — Web UI + chat workstreams + LLM interaction
   # -------------------------------------------------------------------
   server:
-    build:
-      context: .
-      dockerfile: Dockerfile
+    image: turnstone:local
     profiles:
       - production
     command:
@@ -115,6 +113,7 @@ services:
   # turnstone-console — Cluster dashboard
   # -------------------------------------------------------------------
   console:
+    image: turnstone:local
     build:
       context: .
       dockerfile: Dockerfile
@@ -145,9 +144,7 @@ services:
   # Requires TURNSTONE_DISCORD_TOKEN to enable Discord adapter
   # -------------------------------------------------------------------
   channel:
-    build:
-      context: .
-      dockerfile: Dockerfile
+    image: turnstone:local
     profiles:
       - production
       - cluster

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "turnstone"
-version = "1.0.0"
+version = "1.1.0a3"
 description = "Multi-node AI orchestration platform with tool use, agent routing, and cluster simulation."
 readme = "README.md"
 license = "BUSL-1.1"
@@ -67,6 +67,7 @@ turnstone-bootstrap = "turnstone.bootstrap:main"
 [tool.hatch.build.targets.wheel]
 include = [
   "turnstone/**/*.py",
+  "turnstone/prompts/**/*.md",
   "turnstone/tools/*.json",
   "turnstone/ui/static/*.html",
   "turnstone/ui/static/*.css",

sdk/typescript/package-lock.json

@@ -14,22 +14,22 @@
       }
     },
     "node_modules/@emnapi/core": {
-      "version": "1.9.1",
-      "resolved": "https://registry.npmjs.org/@emnapi/core/-/core-1.9.1.tgz",
-      "integrity": "sha512-mukuNALVsoix/w1BJwFzwXBN/dHeejQtuVzcDsfOEsdpCumXb/E9j8w11h5S54tT1xhifGfbbSm/ICrObRb3KA==",
+      "version": "1.9.2",
+      "resolved": "https://registry.npmjs.org/@emnapi/core/-/core-1.9.2.tgz",
+      "integrity": "sha512-UC+ZhH3XtczQYfOlu3lNEkdW/p4dsJ1r/bP7H8+rhao3TTTMO1ATq/4DdIi23XuGoFY+Cz0JmCbdVl0hz9jZcA==",
       "dev": true,
       "license": "MIT",
       "optional": true,
       "peer": true,
       "dependencies": {
-        "@emnapi/wasi-threads": "1.2.0",
+        "@emnapi/wasi-threads": "1.2.1",
         "tslib": "^2.4.0"
       }
     },
     "node_modules/@emnapi/runtime": {
-      "version": "1.9.1",
-      "resolved": "https://registry.npmjs.org/@emnapi/runtime/-/runtime-1.9.1.tgz",
-      "integrity": "sha512-VYi5+ZVLhpgK4hQ0TAjiQiZ6ol0oe4mBx7mVv7IflsiEp0OWoVsp/+f9Vc1hOhE0TtkORVrI1GvzyreqpgWtkA==",
+      "version": "1.9.2",
+      "resolved": "https://registry.npmjs.org/@emnapi/runtime/-/runtime-1.9.2.tgz",
+      "integrity": "sha512-3U4+MIWHImeyu1wnmVygh5WlgfYDtyf0k8AbLhMFxOipihf6nrWC4syIm/SwEeec0mNSafiiNnMJwbza/Is6Lw==",
       "dev": true,
       "license": "MIT",
       "optional": true,
@@ -39,9 +39,9 @@
       }
     },
     "node_modules/@emnapi/wasi-threads": {
-      "version": "1.2.0",
-      "resolved": "https://registry.npmjs.org/@emnapi/wasi-threads/-/wasi-threads-1.2.0.tgz",
-      "integrity": "sha512-N10dEJNSsUx41Z6pZsXU8FjPjpBEplgH24sfkmITrBED1/U2Esum9F3lfLrMjKHHjmi557zQn7kR9R+XWXu5Rg==",
+      "version": "1.2.1",
+      "resolved": "https://registry.npmjs.org/@emnapi/wasi-threads/-/wasi-threads-1.2.1.tgz",
+      "integrity": "sha512-uTII7OYF+/Mes/MrcIOYp5yOtSMLBWSIoLPpcgwipoiKbli6k322tcoFsxoIIxPDqW01SQGAgko4EzZi2BNv2w==",
       "dev": true,
       "license": "MIT",
       "optional": true,

tests/test_hash_ring_storage.py

@@ -41,6 +41,21 @@ class TestHashRingBuckets:
         # Empty list returns 0
         assert storage.assign_buckets([], "node-x") == 0
 
+    def test_assign_large_list_exceeds_chunk_size(self, storage):
+        """Regression: lists larger than chunk_size must not hit param limits."""
+        n = 1200  # exceeds SQLite chunk_size (500) and exercises multi-chunk path
+        storage.seed_ring_buckets([(i, "node-a") for i in range(n)])
+        count = storage.assign_buckets(list(range(n)), "node-b")
+        assert count == n
+        rows = storage.list_ring_buckets()
+        assert all(r["node_id"] == "node-b" for r in rows)
+
+    def test_assign_deduplicates_input(self, storage):
+        """Duplicates in the input list should not inflate rowcount."""
+        storage.seed_ring_buckets([(0, "node-a"), (1, "node-a")])
+        count = storage.assign_buckets([0, 1, 0, 1, 0], "node-b")
+        assert count == 2
+
 
 class TestBucketStats:
     def test_increment_creates_row(self, storage):

tests/test_html.py

@@ -37,3 +37,55 @@ class TestStripHtml:
     def test_self_closing_tags(self):
         result = strip_html("hello<br/>world")
         assert result == "helloworld"
+
+    # -- invisible element stripping -----------------------------------------
+
+    def test_strips_script_content(self):
+        html = "<p>before</p><script>var x = 1;</script><p>after</p>"
+        result = strip_html(html)
+        assert "var x" not in result
+        assert "before" in result
+        assert "after" in result
+
+    def test_strips_style_content(self):
+        html = "<style>.foo { color: red; }</style><p>visible</p>"
+        result = strip_html(html)
+        assert "color" not in result
+        assert "visible" in result
+
+    def test_strips_template_content(self):
+        html = "<template><div>hidden</div></template><p>shown</p>"
+        result = strip_html(html)
+        assert "hidden" not in result
+        assert "shown" in result
+
+    def test_strips_noscript_content(self):
+        html = "<noscript>Enable JS</noscript><p>content</p>"
+        result = strip_html(html)
+        assert "Enable JS" not in result
+        assert "content" in result
+
+    def test_strips_multiple_script_blocks(self):
+        html = "<script>a()</script><p>middle</p><script>b()</script>"
+        result = strip_html(html)
+        assert "a()" not in result
+        assert "b()" not in result
+        assert "middle" in result
+
+    def test_strips_multiline_script(self):
+        html = "<script>\nfunction foo() {\n  return 1;\n}\n</script><p>ok</p>"
+        result = strip_html(html)
+        assert "function" not in result
+        assert "ok" in result
+
+    def test_strips_script_case_insensitive(self):
+        html = "<SCRIPT>code()</SCRIPT><p>text</p>"
+        result = strip_html(html)
+        assert "code()" not in result
+        assert "text" in result
+
+    def test_strips_script_with_attributes(self):
+        html = '<script type="text/javascript" src="app.js">init();</script><p>done</p>'
+        result = strip_html(html)
+        assert "init()" not in result
+        assert "done" in result

tests/test_session.py

@@ -759,6 +759,8 @@ class TestTitleRetry:
     """_generate_title resets _title_generated on failure."""
 
     def test_title_generated_reset_on_failure(self, tmp_db):
+        from turnstone.core.providers._protocol import ModelCapabilities
+
         session = _make_session()
         session._title_generated = True
         session.messages = [
@@ -767,6 +769,7 @@ class TestTitleRetry:
         ]
         # Mock provider to raise
         session._provider = MagicMock()
+        session._provider.get_capabilities.return_value = ModelCapabilities()
         session._provider.create_completion.side_effect = RuntimeError("API error")
 
         session._generate_title()
@@ -774,6 +777,8 @@ class TestTitleRetry:
         assert session._title_generated is False
 
     def test_title_generated_stays_true_on_success(self, tmp_db):
+        from turnstone.core.providers._protocol import ModelCapabilities
+
         session = _make_session()
         session._title_generated = True
         session.messages = [
@@ -783,6 +788,7 @@ class TestTitleRetry:
         result = MagicMock()
         result.content = "Test Title"
         session._provider = MagicMock()
+        session._provider.get_capabilities.return_value = ModelCapabilities()
         session._provider.create_completion.return_value = result
 
         with patch("turnstone.core.session.update_workstream_title"):
@@ -793,6 +799,8 @@ class TestTitleRetry:
 
     def test_title_skipped_after_resume_changes_ws_id(self, tmp_db):
         """If ws_id changes (via resume) during title generation, discard the result."""
+        from turnstone.core.providers._protocol import ModelCapabilities
+
         session = _make_session()
         session._title_generated = True
         session.messages = [
@@ -803,6 +811,7 @@ class TestTitleRetry:
         result = MagicMock()
         result.content = "Test Title"
         session._provider = MagicMock()
+        session._provider.get_capabilities.return_value = ModelCapabilities()
         session._provider.create_completion.return_value = result
 
         # Simulate resume() changing ws_id while title generation is in flight

tests/test_storage_registry.py

@@ -1,8 +1,17 @@
 """Tests for the storage backend registry."""
 
-import pytest
+from unittest.mock import patch
 
-from turnstone.core.storage import get_storage, init_storage, reset_storage
+import pytest
+import sqlalchemy as sa
+
+from turnstone.core.storage import (
+    StorageUnavailableError,
+    get_storage,
+    init_storage,
+    reset_storage,
+)
+from turnstone.core.storage._postgresql import PostgreSQLBackend
 from turnstone.core.storage._sqlite import SQLiteBackend
 
 
@@ -52,3 +61,69 @@ class TestResetStorage:
         init_storage("sqlite", path=str(tmp_path / "test2.db"), run_migrations=False)
         s2 = get_storage()
         assert s1 is not s2
+
+
+class TestConnUnavailableLogging:
+    """Test that _conn() deduplicates DB unavailable/restored logging."""
+
+    def _make_backend(self, tmp_path):
+        """Create a minimal SQLite backend for testing _conn()."""
+        from turnstone.core.storage._sqlite import SQLiteBackend
+
+        return SQLiteBackend(str(tmp_path / "test.db"), create_tables=True)
+
+    def test_logs_unavailable_once(self, tmp_path, caplog: pytest.LogCaptureFixture) -> None:
+        backend = self._make_backend(tmp_path)
+        with patch.object(backend, "_engine") as mock_engine:
+            mock_engine.connect.side_effect = sa.exc.OperationalError(
+                "conn", {}, Exception("refused")
+            )
+            for _ in range(3):
+                with pytest.raises(StorageUnavailableError), backend._conn():
+                    pass  # pragma: no cover
+        unavailable_msgs = [r for r in caplog.records if "database.unavailable" in r.message]
+        assert len(unavailable_msgs) == 1
+
+    def test_logs_restored_on_recovery(self, tmp_path, caplog: pytest.LogCaptureFixture) -> None:
+        import logging
+
+        caplog.set_level(logging.INFO)
+        backend = self._make_backend(tmp_path)
+        # Simulate outage
+        with patch.object(backend, "_engine") as mock_engine:
+            mock_engine.connect.side_effect = sa.exc.OperationalError(
+                "conn", {}, Exception("refused")
+            )
+            with pytest.raises(StorageUnavailableError), backend._conn():
+                pass  # pragma: no cover
+        assert backend._db_unavailable is True
+        # Real connection — should log restored
+        caplog.clear()
+        with backend._conn():
+            pass
+        restored_msgs = [r for r in caplog.records if "database.connection_restored" in r.message]
+        assert len(restored_msgs) == 1
+        assert backend._db_unavailable is False
+
+    def test_postgresql_conn_raises_storage_unavailable(self) -> None:
+        import threading
+
+        backend = PostgreSQLBackend.__new__(PostgreSQLBackend)
+        backend._db_unavailable = False
+        backend._db_unavailable_lock = threading.Lock()
+
+        def _raise_op_error():
+            raise sa.exc.OperationalError("conn", {}, Exception("refused"))
+
+        mock_engine = type(
+            "E",
+            (),
+            {
+                "connect": staticmethod(_raise_op_error),
+                "url": sa.engine.make_url("postgresql://user:pass@localhost/db"),
+            },
+        )()
+        backend._engine = mock_engine
+        with pytest.raises(StorageUnavailableError), backend._conn():
+            pass  # pragma: no cover
+        assert backend._db_unavailable is True

turnstone/__init__.py

@@ -1,3 +1,3 @@
 """turnstone - Multi-node AI orchestration platform with tool use, agent routing, and cluster simulation."""
 
-__version__ = "1.0.0"
+__version__ = "1.1.0a3"

turnstone/channels/cli.py

@@ -282,10 +282,14 @@ def main() -> None:
 
             async def _heartbeat_loop() -> None:
                 """Periodically update service heartbeat."""
+                from turnstone.core.storage._registry import StorageUnavailableError
+
                 while True:
                     await asyncio.sleep(30)
                     try:
                         await asyncio.to_thread(storage.heartbeat_service, "channel", service_id)
+                    except StorageUnavailableError:
+                        pass  # already logged by storage layer
                     except Exception:
                         log.exception("channel.heartbeat_failed")
 

turnstone/console/collector.py

@@ -289,9 +289,13 @@ class ClusterCollector:
 
     def _discovery_loop(self) -> None:
         """Periodically scan the service registry for active nodes."""
+        from turnstone.core.storage._registry import StorageUnavailableError
+
         while self._running:
             try:
                 self._discover_nodes()
+            except StorageUnavailableError:
+                pass  # already logged by storage layer
             except Exception:
                 log.exception("Node discovery error")
             time.sleep(self._discovery_interval)

turnstone/console/rebalancer.py

@@ -19,6 +19,7 @@ from typing import TYPE_CHECKING, Any
 import structlog
 
 from turnstone.core.hash_ring import RING_SIZE, RingNode, bucket_of
+from turnstone.core.storage._registry import StorageUnavailableError
 
 if TYPE_CHECKING:
     from turnstone.console.collector import ClusterCollector
@@ -149,6 +150,8 @@ class Rebalancer:
                 result = self.rebalance_once(trigger=trigger)
                 self._last_result = result
                 self._record_result_metrics(result)
+            except StorageUnavailableError:
+                pass  # already logged by storage layer
             except Exception:
                 log.exception("rebalancer.error")
             finally:

turnstone/console/scheduler.py

@@ -90,9 +90,13 @@ class TaskScheduler:
 
     def _loop(self) -> None:
         """Main scheduler loop — tick then sleep."""
+        from turnstone.core.storage._registry import StorageUnavailableError
+
         while not self._stop_event.is_set():
             try:
                 self._tick()
+            except StorageUnavailableError:
+                pass  # already logged by storage layer
             except Exception:
                 log.exception("scheduler.tick_error")
             self._stop_event.wait(self._check_interval)

turnstone/console/server.py

@@ -1167,10 +1167,14 @@ async def _lifespan(app: Starlette) -> AsyncGenerator[None, None]:
             import asyncio
 
             async def _console_heartbeat() -> None:
+                from turnstone.core.storage._registry import StorageUnavailableError
+
                 while True:
                     await asyncio.sleep(30)
                     try:
                         storage.heartbeat_service("console", "console")
+                    except StorageUnavailableError:
+                        pass  # already logged by storage layer
                     except Exception:
                         log.warning("console.heartbeat_failed", exc_info=True)
 

turnstone/core/providers/_anthropic.py

@@ -543,9 +543,13 @@ class AnthropicProvider:
         if extra_params and "thinking_budget_tokens" in extra_params:
             budget = extra_params["thinking_budget_tokens"]
         if budget > 0:
-            # Budget must leave room for the response
+            # Budget must be strictly less than max_tokens (API requirement).
+            # If max_tokens is too small to fit even a minimal thinking
+            # budget alongside the response, disable thinking entirely.
             if budget >= max_tokens:
-                budget = max(1024, max_tokens - 1024)
+                budget = max_tokens - 1024
+            if budget < 1:
+                return {}
             return {"thinking": {"type": "enabled", "budget_tokens": budget}}
         return {}
 

turnstone/core/session.py

@@ -924,10 +924,8 @@ class ChatSession:
             if asst_msg:
                 snippet += f"\nAssistant: {asst_msg}"
             snippet += "\n\nTitle:"
-            result = self._provider.create_completion(
-                client=self.client,
-                model=self.model,
-                messages=[
+            result = self._utility_completion(
+                [
                     {
                         "role": "system",
                         "content": (
@@ -942,9 +940,6 @@ class ChatSession:
                     {"role": "user", "content": snippet},
                 ],
                 max_tokens=200,
-                temperature=0.3,
-                reasoning_effort="low",
-                extra_params=self._provider_extra_params(reasoning_effort="low"),
             )
             raw = (result.content or "").strip()
             # Take first line, strip quotes
@@ -1281,6 +1276,33 @@ class ChatSession:
             return {"chat_template_kwargs": kwargs}
         return None
 
+    def _utility_completion(
+        self,
+        messages: list[dict[str, Any]],
+        *,
+        max_tokens: int = 4096,
+        temperature: float = 0.3,
+        reasoning_effort: str = "low",
+    ) -> CompletionResult:
+        """Run a lightweight internal completion (title gen, compaction, extraction).
+
+        Threads ``reasoning_effort`` through both the direct keyword (for
+        commercial providers) and ``extra_params`` (for local model servers)
+        so callers don't need to duplicate it.  ``max_tokens`` is clamped to
+        the model's advertised output limit so small models don't error.
+        """
+        caps = self._get_capabilities()
+        clamped = min(max_tokens, caps.max_output_tokens) if caps.max_output_tokens else max_tokens
+        return self._provider.create_completion(
+            client=self.client,
+            model=self.model,
+            messages=messages,
+            max_tokens=clamped,
+            temperature=temperature,
+            reasoning_effort=reasoning_effort,
+            extra_params=self._provider_extra_params(reasoning_effort=reasoning_effort),
+        )
+
     # -- tool search helpers --------------------------------------------------
 
     def _get_active_tools(self) -> list[dict[str, Any]] | None:
@@ -2482,14 +2504,9 @@ class ChatSession:
             result: CompletionResult | None = None
             for attempt in range(self._MAX_RETRIES + 1):
                 try:
-                    result = self._provider.create_completion(
-                        client=self.client,
-                        model=self.model,
-                        messages=summary_msgs,
+                    result = self._utility_completion(
+                        summary_msgs,
                         max_tokens=summary_max_tokens,
-                        temperature=0.3,
-                        reasoning_effort="low",
-                        extra_params=self._provider_extra_params(reasoning_effort="low"),
                     )
                     break
                 except Exception as e:
@@ -6167,26 +6184,30 @@ class ChatSession:
             return call_id, msg
 
         if not text.strip():
-            return call_id, "(empty response from URL)"
+            msg = "Error: fetch returned empty response"
+            self._report_tool_result(call_id, "web_fetch", msg, is_error=True)
+            return call_id, msg
 
         original_len = len(text)
         self.ui.on_info(f"fetched {original_len} chars, extracting...")
 
-        # Phase 2: truncate for summarization context
-        max_content = 50_000
+        # Phase 2: truncate for summarization context.
+        # Reserve ~25% of the context window for the extraction prompt
+        # overhead (system message, URL, question) and response tokens.
+        # Convert token budget to chars using the calibrated ratio.
+        max_content = int(self.context_window * self._chars_per_token * 0.75)
+        max_content = min(max(max_content, 50_000), 500_000)  # 50k–500k
         if len(text) > max_content:
-            text = (
-                text[: max_content // 2]
-                + f"\n\n... [{len(text) - max_content} chars omitted] ...\n\n"
-                + text[-(max_content // 2) :]
-            )
+            # Prefer the beginning — page content is usually top-heavy.
+            text = text[:max_content] + f"\n\n... [{len(text) - max_content} chars truncated] ...\n"
 
-        # Phase 3: summarization API call
+        # Phase 3: summarization API call.
+        # Use a generous max_tokens so thinking models don't starve the
+        # visible answer, and pass reasoning_effort="low" to avoid wasting
+        # budget on deep reasoning for a simple extraction task.
         try:
-            result = self._provider.create_completion(
-                client=self.client,
-                model=self.model,
-                messages=[
+            result = self._utility_completion(
+                [
                     {
                         "role": "system",
                         "content": (
@@ -6206,11 +6227,12 @@ class ChatSession:
                         ),
                     },
                 ],
-                max_tokens=2000,
+                max_tokens=8192,
                 temperature=0.2,
-                extra_params=self._provider_extra_params(),
             )
-            answer = result.content or "(no answer)"
+            answer = result.content or ""
+            if not answer:
+                answer = "Error: extraction returned no answer"
         except Exception as e:
             answer = f"Extraction failed (page was fetched but summarization errored): {e}"
 
@@ -6218,7 +6240,7 @@ class ChatSession:
             call_id,
             "web_fetch",
             answer,
-            is_error=answer.startswith("Extraction failed"),
+            is_error=answer.startswith(("Error:", "Extraction failed")),
         )
 
         return call_id, answer

turnstone/core/storage/__init__.py

@@ -4,10 +4,16 @@ Supports SQLite (default, zero-config) and PostgreSQL (multi-node, production).
 """
 
 from turnstone.core.storage._protocol import StorageBackend
-from turnstone.core.storage._registry import get_storage, init_storage, reset_storage
+from turnstone.core.storage._registry import (
+    StorageUnavailableError,
+    get_storage,
+    init_storage,
+    reset_storage,
+)
 
 __all__ = [
     "StorageBackend",
+    "StorageUnavailableError",
     "get_storage",
     "init_storage",
     "reset_storage",

turnstone/core/storage/_registry.py

@@ -15,6 +15,14 @@ log = get_logger(__name__)
 _storage: StorageBackend | None = None
 
 
+class StorageUnavailableError(Exception):
+    """Raised when the database is unreachable.
+
+    The storage layer has already logged a clean one-liner — callers
+    should catch this to avoid duplicate tracebacks.
+    """
+
+
 def init_storage(
     backend: str = "sqlite",
     *,

turnstone/core/watch.py

@@ -271,9 +271,13 @@ class WatchRunner:
     # -- Main loop -----------------------------------------------------------
 
     def _run(self) -> None:
+        from turnstone.core.storage._registry import StorageUnavailableError
+
         while not self._stop_event.is_set():
             try:
                 self._tick()
+            except StorageUnavailableError:
+                pass  # already logged by storage layer
             except Exception:
                 log.exception("watch_runner.tick_error")
             self._stop_event.wait(self._check_interval)

turnstone/core/web.py

@@ -6,14 +6,20 @@ import socket
 from html import unescape as _html_unescape
 from urllib.parse import urlparse
 
+_RE_INVISIBLE = re.compile(
+    r"<(script|style|template|noscript)\b[^>]*>.*?</\1\s*>",
+    re.DOTALL | re.IGNORECASE,
+)
 _RE_TAGS = re.compile(r"<[^>]+>")
 _RE_WS = re.compile(r"[ \t]+")
 _RE_BLANKLINES = re.compile(r"\n{3,}")
 
 
 def strip_html(html: str) -> str:
-    """Convert HTML to plain text: strip tags, decode entities, collapse whitespace."""
-    text = _RE_TAGS.sub("", html)
+    """Convert HTML to plain text: strip invisible elements, tags, decode entities."""
+    # Remove elements whose content should never appear as text
+    text = _RE_INVISIBLE.sub("", html)
+    text = _RE_TAGS.sub("", text)
     text = _html_unescape(text)
     text = _RE_WS.sub(" ", text)
     text = _RE_BLANKLINES.sub("\n\n", text)

turnstone/server.py

@@ -2389,10 +2389,14 @@ async def _lifespan(app: Starlette) -> AsyncGenerator[None, None]:
 
         async def _heartbeat_loop() -> None:
             """Periodically update service heartbeat."""
+            from turnstone.core.storage._registry import StorageUnavailableError
+
             while True:
                 await asyncio.sleep(30)
                 try:
                     await asyncio.to_thread(_svc_storage.heartbeat_service, "server", _svc_node_id)
+                except StorageUnavailableError:
+                    pass  # already logged by storage layer
                 except Exception:
                     log.exception("server.heartbeat_failed")
 

uv.lock

@@ -1249,7 +1249,7 @@ wheels = [
 
 [[package]]
 name = "mcp"
-version = "1.26.0"
+version = "1.27.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "anyio" },
@@ -1267,9 +1267,9 @@ dependencies = [
     { name = "typing-inspection" },
     { name = "uvicorn", marker = "sys_platform != 'emscripten'" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/fc/6d/62e76bbb8144d6ed86e202b5edd8a4cb631e7c8130f3f4893c3f90262b10/mcp-1.26.0.tar.gz", hash = "sha256:db6e2ef491eecc1a0d93711a76f28dec2e05999f93afd48795da1c1137142c66", size = 608005, upload-time = "2026-01-24T19:40:32.468Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/8b/eb/c0cfc62075dc6e1ec1c64d352ae09ac051d9334311ed226f1f425312848a/mcp-1.27.0.tar.gz", hash = "sha256:d3dc35a7eec0d458c1da4976a48f982097ddaab87e278c5511d5a4a56e852b83", size = 607509, upload-time = "2026-04-02T14:48:08.88Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/fd/d9/eaa1f80170d2b7c5ba23f3b59f766f3a0bb41155fbc32a69adfa1adaaef9/mcp-1.26.0-py3-none-any.whl", hash = "sha256:904a21c33c25aa98ddbeb47273033c435e595bbacfdb177f4bd87f6dceebe1ca", size = 233615, upload-time = "2026-01-24T19:40:30.652Z" },
+    { url = "https://files.pythonhosted.org/packages/9c/46/f6b4ad632c67ef35209a66127e4bddc95759649dd595f71f13fba11bdf9a/mcp-1.27.0-py3-none-any.whl", hash = "sha256:5ce1fa81614958e267b21fb2aa34e0aea8e2c6ede60d52aba45fd47246b4d741", size = 215967, upload-time = "2026-04-02T14:48:07.24Z" },
 ]
 
 [[package]]
@@ -2216,27 +2216,27 @@ wheels = [
 
 [[package]]
 name = "ruff"
-version = "0.15.8"
+version = "0.15.9"
 source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/14/b0/73cf7550861e2b4824950b8b52eebdcc5adc792a00c514406556c5b80817/ruff-0.15.8.tar.gz", hash = "sha256:995f11f63597ee362130d1d5a327a87cb6f3f5eae3094c620bcc632329a4d26e", size = 4610921, upload-time = "2026-03-26T18:39:38.675Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/e6/97/e9f1ca355108ef7194e38c812ef40ba98c7208f47b13ad78d023caa583da/ruff-0.15.9.tar.gz", hash = "sha256:29cbb1255a9797903f6dde5ba0188c707907ff44a9006eb273b5a17bfa0739a2", size = 4617361, upload-time = "2026-04-02T18:17:20.829Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/4a/92/c445b0cd6da6e7ae51e954939cb69f97e008dbe750cfca89b8cedc081be7/ruff-0.15.8-py3-none-linux_armv6l.whl", hash = "sha256:cbe05adeba76d58162762d6b239c9056f1a15a55bd4b346cfd21e26cd6ad7bc7", size = 10527394, upload-time = "2026-03-26T18:39:41.566Z" },
-    { url = "https://files.pythonhosted.org/packages/eb/92/f1c662784d149ad1414cae450b082cf736430c12ca78367f20f5ed569d65/ruff-0.15.8-py3-none-macosx_10_12_x86_64.whl", hash = "sha256:d3e3d0b6ba8dca1b7ef9ab80a28e840a20070c4b62e56d675c24f366ef330570", size = 10905693, upload-time = "2026-03-26T18:39:30.364Z" },
-    { url = "https://files.pythonhosted.org/packages/ca/f2/7a631a8af6d88bcef997eb1bf87cc3da158294c57044aafd3e17030613de/ruff-0.15.8-py3-none-macosx_11_0_arm64.whl", hash = "sha256:6ee3ae5c65a42f273f126686353f2e08ff29927b7b7e203b711514370d500de3", size = 10323044, upload-time = "2026-03-26T18:39:33.37Z" },
-    { url = "https://files.pythonhosted.org/packages/67/18/1bf38e20914a05e72ef3b9569b1d5c70a7ef26cd188d69e9ca8ef588d5bf/ruff-0.15.8-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:fdce027ada77baa448077ccc6ebb2fa9c3c62fd110d8659d601cf2f475858d94", size = 10629135, upload-time = "2026-03-26T18:39:44.142Z" },
-    { url = "https://files.pythonhosted.org/packages/d2/e9/138c150ff9af60556121623d41aba18b7b57d95ac032e177b6a53789d279/ruff-0.15.8-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:12e617fc01a95e5821648a6df341d80456bd627bfab8a829f7cfc26a14a4b4a3", size = 10348041, upload-time = "2026-03-26T18:39:52.178Z" },
-    { url = "https://files.pythonhosted.org/packages/02/f1/5bfb9298d9c323f842c5ddeb85f1f10ef51516ac7a34ba446c9347d898df/ruff-0.15.8-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:432701303b26416d22ba696c39f2c6f12499b89093b61360abc34bcc9bf07762", size = 11121987, upload-time = "2026-03-26T18:39:55.195Z" },
-    { url = "https://files.pythonhosted.org/packages/10/11/6da2e538704e753c04e8d86b1fc55712fdbdcc266af1a1ece7a51fff0d10/ruff-0.15.8-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:d910ae974b7a06a33a057cb87d2a10792a3b2b3b35e33d2699fdf63ec8f6b17a", size = 11951057, upload-time = "2026-03-26T18:39:19.18Z" },
-    { url = "https://files.pythonhosted.org/packages/83/f0/c9208c5fd5101bf87002fed774ff25a96eea313d305f1e5d5744698dc314/ruff-0.15.8-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:2033f963c43949d51e6fdccd3946633c6b37c484f5f98c3035f49c27395a8ab8", size = 11464613, upload-time = "2026-03-26T18:40:06.301Z" },
-    { url = "https://files.pythonhosted.org/packages/f8/22/d7f2fabdba4fae9f3b570e5605d5eb4500dcb7b770d3217dca4428484b17/ruff-0.15.8-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0f29b989a55572fb885b77464cf24af05500806ab4edf9a0fd8977f9759d85b1", size = 11257557, upload-time = "2026-03-26T18:39:57.972Z" },
-    { url = "https://files.pythonhosted.org/packages/71/8c/382a9620038cf6906446b23ce8632ab8c0811b8f9d3e764f58bedd0c9a6f/ruff-0.15.8-py3-none-manylinux_2_31_riscv64.whl", hash = "sha256:ac51d486bf457cdc985a412fb1801b2dfd1bd8838372fc55de64b1510eff4bec", size = 11169440, upload-time = "2026-03-26T18:39:22.205Z" },
-    { url = "https://files.pythonhosted.org/packages/4d/0d/0994c802a7eaaf99380085e4e40c845f8e32a562e20a38ec06174b52ef24/ruff-0.15.8-py3-none-musllinux_1_2_aarch64.whl", hash = "sha256:c9861eb959edab053c10ad62c278835ee69ca527b6dcd72b47d5c1e5648964f6", size = 10605963, upload-time = "2026-03-26T18:39:46.682Z" },
-    { url = "https://files.pythonhosted.org/packages/19/aa/d624b86f5b0aad7cef6bbf9cd47a6a02dfdc4f72c92a337d724e39c9d14b/ruff-0.15.8-py3-none-musllinux_1_2_armv7l.whl", hash = "sha256:8d9a5b8ea13f26ae90838afc33f91b547e61b794865374f114f349e9036835fb", size = 10357484, upload-time = "2026-03-26T18:39:49.176Z" },
-    { url = "https://files.pythonhosted.org/packages/35/c3/e0b7835d23001f7d999f3895c6b569927c4d39912286897f625736e1fd04/ruff-0.15.8-py3-none-musllinux_1_2_i686.whl", hash = "sha256:c2a33a529fb3cbc23a7124b5c6ff121e4d6228029cba374777bd7649cc8598b8", size = 10830426, upload-time = "2026-03-26T18:40:03.702Z" },
-    { url = "https://files.pythonhosted.org/packages/f0/51/ab20b322f637b369383adc341d761eaaa0f0203d6b9a7421cd6e783d81b9/ruff-0.15.8-py3-none-musllinux_1_2_x86_64.whl", hash = "sha256:75e5cd06b1cf3f47a3996cfc999226b19aa92e7cce682dcd62f80d7035f98f49", size = 11345125, upload-time = "2026-03-26T18:39:27.799Z" },
-    { url = "https://files.pythonhosted.org/packages/37/e6/90b2b33419f59d0f2c4c8a48a4b74b460709a557e8e0064cf33ad894f983/ruff-0.15.8-py3-none-win32.whl", hash = "sha256:bc1f0a51254ba21767bfa9a8b5013ca8149dcf38092e6a9eb704d876de94dc34", size = 10571959, upload-time = "2026-03-26T18:39:36.117Z" },
-    { url = "https://files.pythonhosted.org/packages/1f/a2/ef467cb77099062317154c63f234b8a7baf7cb690b99af760c5b68b9ee7f/ruff-0.15.8-py3-none-win_amd64.whl", hash = "sha256:04f79eff02a72db209d47d665ba7ebcad609d8918a134f86cb13dd132159fc89", size = 11743893, upload-time = "2026-03-26T18:39:25.01Z" },
-    { url = "https://files.pythonhosted.org/packages/15/e2/77be4fff062fa78d9b2a4dea85d14785dac5f1d0c1fb58ed52331f0ebe28/ruff-0.15.8-py3-none-win_arm64.whl", hash = "sha256:cf891fa8e3bb430c0e7fac93851a5978fc99c8fa2c053b57b118972866f8e5f2", size = 11048175, upload-time = "2026-03-26T18:40:01.06Z" },
+    { url = "https://files.pythonhosted.org/packages/0b/1f/9cdfd0ac4b9d1e5a6cf09bedabdf0b56306ab5e333c85c87281273e7b041/ruff-0.15.9-py3-none-linux_armv6l.whl", hash = "sha256:6efbe303983441c51975c243e26dff328aca11f94b70992f35b093c2e71801e1", size = 10511206, upload-time = "2026-04-02T18:16:41.574Z" },
+    { url = "https://files.pythonhosted.org/packages/3d/f6/32bfe3e9c136b35f02e489778d94384118bb80fd92c6d92e7ccd97db12ce/ruff-0.15.9-py3-none-macosx_10_12_x86_64.whl", hash = "sha256:4965bac6ac9ea86772f4e23587746f0b7a395eccabb823eb8bfacc3fa06069f7", size = 10923307, upload-time = "2026-04-02T18:17:08.645Z" },
+    { url = "https://files.pythonhosted.org/packages/ca/25/de55f52ab5535d12e7aaba1de37a84be6179fb20bddcbe71ec091b4a3243/ruff-0.15.9-py3-none-macosx_11_0_arm64.whl", hash = "sha256:eaf05aad70ca5b5a0a4b0e080df3a6b699803916d88f006efd1f5b46302daab8", size = 10316722, upload-time = "2026-04-02T18:16:44.206Z" },
+    { url = "https://files.pythonhosted.org/packages/48/11/690d75f3fd6278fe55fff7c9eb429c92d207e14b25d1cae4064a32677029/ruff-0.15.9-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:9439a342adb8725f32f92732e2bafb6d5246bd7a5021101166b223d312e8fc59", size = 10623674, upload-time = "2026-04-02T18:16:50.951Z" },
+    { url = "https://files.pythonhosted.org/packages/bd/ec/176f6987be248fc5404199255522f57af1b4a5a1b57727e942479fec98ad/ruff-0.15.9-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:9c5e6faf9d97c8edc43877c3f406f47446fc48c40e1442d58cfcdaba2acea745", size = 10351516, upload-time = "2026-04-02T18:16:57.206Z" },
+    { url = "https://files.pythonhosted.org/packages/b2/fc/51cffbd2b3f240accc380171d51446a32aa2ea43a40d4a45ada67368fbd2/ruff-0.15.9-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:7b34a9766aeec27a222373d0b055722900fbc0582b24f39661aa96f3fe6ad901", size = 11150202, upload-time = "2026-04-02T18:17:06.452Z" },
+    { url = "https://files.pythonhosted.org/packages/d6/d4/25292a6dfc125f6b6528fe6af31f5e996e19bf73ca8e3ce6eb7fa5b95885/ruff-0.15.9-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:89dd695bc72ae76ff484ae54b7e8b0f6b50f49046e198355e44ea656e521fef9", size = 11988891, upload-time = "2026-04-02T18:17:18.575Z" },
+    { url = "https://files.pythonhosted.org/packages/13/e1/1eebcb885c10e19f969dcb93d8413dfee8172578709d7ee933640f5e7147/ruff-0.15.9-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ce187224ef1de1bd225bc9a152ac7102a6171107f026e81f317e4257052916d5", size = 11480576, upload-time = "2026-04-02T18:16:52.986Z" },
+    { url = "https://files.pythonhosted.org/packages/ff/6b/a1548ac378a78332a4c3dcf4a134c2475a36d2a22ddfa272acd574140b50/ruff-0.15.9-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:2b0c7c341f68adb01c488c3b7d4b49aa8ea97409eae6462d860a79cf55f431b6", size = 11254525, upload-time = "2026-04-02T18:17:02.041Z" },
+    { url = "https://files.pythonhosted.org/packages/42/aa/4bb3af8e61acd9b1281db2ab77e8b2c3c5e5599bf2a29d4a942f1c62b8d6/ruff-0.15.9-py3-none-manylinux_2_31_riscv64.whl", hash = "sha256:55cc15eee27dc0eebdfcb0d185a6153420efbedc15eb1d38fe5e685657b0f840", size = 11204072, upload-time = "2026-04-02T18:17:13.581Z" },
+    { url = "https://files.pythonhosted.org/packages/69/48/d550dc2aa6e423ea0bcc1d0ff0699325ffe8a811e2dba156bd80750b86dc/ruff-0.15.9-py3-none-musllinux_1_2_aarch64.whl", hash = "sha256:a6537f6eed5cda688c81073d46ffdfb962a5f29ecb6f7e770b2dc920598997ed", size = 10594998, upload-time = "2026-04-02T18:16:46.369Z" },
+    { url = "https://files.pythonhosted.org/packages/63/47/321167e17f5344ed5ec6b0aa2cff64efef5f9e985af8f5622cfa6536043f/ruff-0.15.9-py3-none-musllinux_1_2_armv7l.whl", hash = "sha256:6d3fcbca7388b066139c523bda744c822258ebdcfbba7d24410c3f454cc9af71", size = 10359769, upload-time = "2026-04-02T18:17:10.994Z" },
+    { url = "https://files.pythonhosted.org/packages/67/5e/074f00b9785d1d2c6f8c22a21e023d0c2c1817838cfca4c8243200a1fa87/ruff-0.15.9-py3-none-musllinux_1_2_i686.whl", hash = "sha256:058d8e99e1bfe79d8a0def0b481c56059ee6716214f7e425d8e737e412d69677", size = 10850236, upload-time = "2026-04-02T18:16:48.749Z" },
+    { url = "https://files.pythonhosted.org/packages/76/37/804c4135a2a2caf042925d30d5f68181bdbd4461fd0d7739da28305df593/ruff-0.15.9-py3-none-musllinux_1_2_x86_64.whl", hash = "sha256:8e1ddb11dbd61d5983fa2d7d6370ef3eb210951e443cace19594c01c72abab4c", size = 11358343, upload-time = "2026-04-02T18:16:55.068Z" },
+    { url = "https://files.pythonhosted.org/packages/88/3d/1364fcde8656962782aa9ea93c92d98682b1ecec2f184e625a965ad3b4a6/ruff-0.15.9-py3-none-win32.whl", hash = "sha256:bde6ff36eaf72b700f32b7196088970bf8fdb2b917b7accd8c371bfc0fd573ec", size = 10583382, upload-time = "2026-04-02T18:17:04.261Z" },
+    { url = "https://files.pythonhosted.org/packages/4c/56/5c7084299bd2cacaa07ae63a91c6f4ba66edc08bf28f356b24f6b717c799/ruff-0.15.9-py3-none-win_amd64.whl", hash = "sha256:45a70921b80e1c10cf0b734ef09421f71b5aa11d27404edc89d7e8a69505e43d", size = 11744969, upload-time = "2026-04-02T18:16:59.611Z" },
+    { url = "https://files.pythonhosted.org/packages/03/36/76704c4f312257d6dbaae3c959add2a622f63fcca9d864659ce6d8d97d3d/ruff-0.15.9-py3-none-win_arm64.whl", hash = "sha256:0694e601c028fd97dc5c6ee244675bc241aeefced7ef80cd9c6935a871078f53", size = 11005870, upload-time = "2026-04-02T18:17:15.773Z" },
 ]
 
 [[package]]
@@ -2496,7 +2496,7 @@ wheels = [
 
 [[package]]
 name = "turnstone"
-version = "1.0.0"
+version = "1.1.0a3"
 source = { editable = "." }
 dependencies = [
     { name = "alembic" },