topos-node 1.0.0__tar.gz → 1.0.2__tar.gz

{topos_node-1.0.0 → topos_node-1.0.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: topos-node
-Version: 1.0.0
+Version: 1.0.2
 Summary: Topos personal AI engine (FastAPI): local data, sync, control plane WebSocket client
 License-Expression: Apache-2.0
 Project-URL: Homepage, https://topos.dialogues.ai
@@ -23,6 +23,16 @@ Requires-Dist: mcp~=1.27.0
 Requires-Dist: google-cloud-storage~=3.10.1
 Requires-Dist: google-cloud-run~=0.16.0
 Requires-Dist: asgi-lifespan~=2.1.0
+Requires-Dist: psycopg[binary]~=3.3.3
+Requires-Dist: duckdb~=1.5.2
+Requires-Dist: torch<3,>=2.8.0
+Requires-Dist: transformers~=5.10.2
+Requires-Dist: huggingface-hub~=1.18.0
+Requires-Dist: tqdm~=4.67.3
+Requires-Dist: hf-xet~=1.5.1
+Requires-Dist: sentence-transformers~=5.6.0
+Requires-Dist: networkx<3.7,>=3.2.1
+Requires-Dist: python-louvain<0.17,>=0.16
 Provides-Extra: database
 Requires-Dist: psycopg[binary]~=3.3.3; extra == "database"
 Provides-Extra: engine
@@ -36,16 +46,6 @@ Requires-Dist: sentence-transformers~=5.6.0; extra == "engine"
 Requires-Dist: networkx<3.7,>=3.2.1; extra == "engine"
 Requires-Dist: python-louvain<0.17,>=0.16; extra == "engine"
 Provides-Extra: local
-Requires-Dist: psycopg[binary]~=3.3.3; extra == "local"
-Requires-Dist: duckdb~=1.5.2; extra == "local"
-Requires-Dist: torch<3,>=2.8.0; extra == "local"
-Requires-Dist: transformers~=5.10.2; extra == "local"
-Requires-Dist: huggingface-hub~=1.18.0; extra == "local"
-Requires-Dist: tqdm~=4.67.3; extra == "local"
-Requires-Dist: hf-xet~=1.5.1; extra == "local"
-Requires-Dist: sentence-transformers~=5.6.0; extra == "local"
-Requires-Dist: networkx<3.7,>=3.2.1; extra == "local"
-Requires-Dist: python-louvain<0.17,>=0.16; extra == "local"
 Provides-Extra: signal
 Requires-Dist: pysqlcipher3~=1.2.0; extra == "signal"
 Provides-Extra: dev
@@ -122,7 +122,7 @@ The `shared/` package in this repo is part of the node runtime contract. It cont
 ### 1) Install
 
 ```bash
-uv tool install 'topos-node[local]'
+uv tool install topos-node
 ```
 
 ### 2) Configure

{topos_node-1.0.0 → topos_node-1.0.2}/README.md

@@ -66,7 +66,7 @@ The `shared/` package in this repo is part of the node runtime contract. It cont
 ### 1) Install
 
 ```bash
-uv tool install 'topos-node[local]'
+uv tool install topos-node
 ```
 
 ### 2) Configure

{topos_node-1.0.0 → topos_node-1.0.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "topos-node"
-version = "1.0.0"
+version = "1.0.2"
 description = "Topos personal AI engine (FastAPI): local data, sync, control plane WebSocket client"
 readme = "README.md"
 license = "Apache-2.0"
@@ -25,6 +25,16 @@ dependencies = [
   "google-cloud-storage~=3.10.1",
   "google-cloud-run~=0.16.0",
   "asgi-lifespan~=2.1.0",
+  "psycopg[binary]~=3.3.3",
+  "duckdb~=1.5.2",
+  "torch>=2.8.0,<3",
+  "transformers~=5.10.2",
+  "huggingface-hub~=1.18.0",
+  "tqdm~=4.67.3",
+  "hf-xet~=1.5.1",
+  "sentence-transformers~=5.6.0",
+  "networkx>=3.2.1,<3.7",
+  "python-louvain>=0.16,<0.17",
 ]
 
 [project.urls]
@@ -48,19 +58,8 @@ engine = [
   "networkx>=3.2.1,<3.7",
   "python-louvain>=0.16,<0.17",
 ]
-# Colocated local node (database + engine on one device). Default for PyPI / uv tool install.
-local = [
-  "psycopg[binary]~=3.3.3",
-  "duckdb~=1.5.2",
-  "torch>=2.8.0,<3",
-  "transformers~=5.10.2",
-  "huggingface-hub~=1.18.0",
-  "tqdm~=4.67.3",
-  "hf-xet~=1.5.1",
-  "sentence-transformers~=5.6.0",
-  "networkx>=3.2.1,<3.7",
-  "python-louvain>=0.16,<0.17",
-]
+# Backward-compatible alias; local runtime deps are in core dependencies.
+local = []
 signal = [
   "pysqlcipher3~=1.2.0",
 ]

{topos_node-1.0.0 → topos_node-1.0.2}/topos/__version__.py

@@ -1,6 +1,6 @@
 """Topos Control Plane Version Information."""
 
-__version__ = "1.0.0"
-__version_info__ = (1, 0, 0)
+__version__ = "1.0.2"
+__version_info__ = (1, 0, 2)
 
 __all__ = ["__version__", "__version_info__"]

{topos_node-1.0.0 → topos_node-1.0.2}/topos/api/enrichment.py

@@ -94,7 +94,7 @@ async def _backfill_browser_visits_url_classification(
     )
     from ..sources.registry import BROWSER_VISITS
 
-    logger.info(
+    logger.debug(
         "[PIPELINE:SIGNAL_DERIVE] Browser visits backfill from activity_events only_missing=%s limit=%s",
         only_missing,
         limit,
@@ -175,7 +175,7 @@ async def _backfill_browser_visits_url_classification(
         "rows_failed": failed,
         "errors": (derive_result.get("errors") or [])[:100],
     }
-    logger.info(
+    logger.debug(
         "[PIPELINE:SIGNAL_DERIVE] Browser visits backfill complete: scanned=%d processed=%d skipped=%d failed=%d",
         summary["rows_scanned"],
         summary["rows_processed"],
@@ -549,7 +549,7 @@ async def _process_enrichment_core(
     tables_manager = DerivedTablesManager(conn=db_conn)
     orchestrator = EnrichmentOrchestrator(tables_manager=tables_manager)
     
-    logger.info(
+    logger.debug(
         "[PIPELINE:ENRICHMENT] %s: Manual enrichment triggered: source_id=%s, messages=%d, jobs=%s",
         orchestrator,
         source_id,

{topos_node-1.0.0 → topos_node-1.0.2}/topos/config/settings.py

@@ -8,6 +8,9 @@ from typing import List, Optional
 from pydantic import Field, model_validator
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
+# Set before any huggingface_hub import during app/route loading.
+os.environ.setdefault("HF_HUB_DISABLE_PROGRESS_BARS", "1")
+
 
 DEFAULT_TOPOS_CONTROL_PLANE_URL = "wss://cp.logu3s.com/ws/engine"
 

{topos_node-1.0.0 → topos_node-1.0.2}/topos/core/handlers.py

@@ -171,7 +171,7 @@ def _uma_transform_progress_hook(req_id: str, stage_label: str):
             return
         state["bucket"] = bucket
         filter_text = f" filter={current_filter}" if current_filter else ""
-        logger.info(
+        logger.debug(
             "[PIPELINE:UMA][TRANSFORM] req=%s stage=%s%s %s %s%% (%s/%s)",
             req_id,
             stage_label,
@@ -2118,7 +2118,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                 if not existing_user_id:
                     # Engine started but no user_id yet - link to authenticated user
                     store_user_id(conn, auth_user_id)
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:CONNECTION] Engine had no user_id. Stored auth user_id=%s from control plane. "
                         "Engine is now linked to authenticated user.",
                         auth_user_id[:8] if auth_user_id else None,
@@ -4178,7 +4178,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
         if not enrichment_name:
             return {"id": req_id, "status": "error", "error": "enrichment_name required"}
         try:
-            logger.info(
+            logger.debug(
                 "[PIPELINE:ENRICHMENT] source_enrichment_backfill received: source_id=%s enrichment=%s only_missing=%s limit=%s",
                 source_id,
                 enrichment_name,
@@ -4193,7 +4193,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                 only_missing=bool(only_missing),
                 limit=limit,
             )
-            logger.info(
+            logger.debug(
                 "[PIPELINE:ENRICHMENT] source_enrichment_backfill complete: source_id=%s enrichment=%s rows_scanned=%s rows_processed=%s rows_failed=%s",
                 source_id,
                 enrichment_name,
@@ -4215,7 +4215,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
         if not enrichment_name:
             return {"id": req_id, "status": "error", "error": "enrichment_name required"}
         try:
-            logger.info(
+            logger.debug(
                 "[PIPELINE:ENRICHMENT] source_enrichment_test received: source_id=%s enrichment=%s",
                 source_id,
                 enrichment_name,
@@ -5359,7 +5359,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
         requesting_app_id = (payload.get("requesting_app_id") or "").strip() or None
         requesting_app_name = (payload.get("requesting_app_name") or "").strip() or None
         app_display = requesting_app_name or requesting_app_id or "(unknown app)"
-        logger.info(
+        logger.debug(
             "[PIPELINE:UMA] uma_get_messages: message_stream=%s, resource_id=%s, dataset_id=%s, limit=%s, offset=%s, requesting_user_email=%s, requesting_app=%s",
             message_stream,
             resource_id[:50] if resource_id else None,
@@ -5502,12 +5502,12 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                         pre_name_rows = sum(1 for row in after_contact if row.get("sender_display_name"))
                         manifest_for_generic = strip_contact_runtime_filters(filter_manifest)
                         transform_diag: Dict[str, Any] = {}
-                        logger.info(
+                        logger.debug(
                             "[PIPELINE:UMA][TRANSFORM] req=%s stage=conversation_messages start rows=%s",
                             req_id,
                             len(after_contact),
                         )
-                        logger.info(
+                        logger.debug(
                             "[PIPELINE:UMA] contact names: contacts_resolve=%s names_effective=%s "
                             "sender_display_name_rows_after_pipeline=%s allowed_scopes=%s",
                             contacts_resolve,
@@ -5523,7 +5523,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                             diagnostics=transform_diag,
                             progress_hook=_uma_transform_progress_hook(req_id, "conversation_messages"),
                         )
-                        logger.info(
+                        logger.debug(
                             "[PIPELINE:UMA][TRANSFORM] req=%s stage=conversation_messages done applied=%s skipped=%s reasons=%s",
                             req_id,
                             transform_diag.get("applied_count", 0),
@@ -5532,7 +5532,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                         )
                         _skip_reasons = transform_diag.get("skip_reasons") or {}
                         if _skip_reasons.get("table_mismatch"):
-                            logger.info(
+                            logger.debug(
                                 "[PIPELINE:UMA] req=%s table_mismatch skips=%s are expected when field_transforms "
                                 "include other tables (browser_visits, etc.) while processing conversation_messages.",
                                 req_id,
@@ -5707,7 +5707,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                         pre_contact_len,
                     )
                 transform_diag = {}
-                logger.info(
+                logger.debug(
                     "[PIPELINE:UMA][TRANSFORM] req=%s stage=ai_chat_messages start rows=%s",
                     req_id,
                     len(messages),
@@ -5720,7 +5720,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                     diagnostics=transform_diag,
                     progress_hook=_uma_transform_progress_hook(req_id, "ai_chat_messages"),
                 )
-                logger.info(
+                logger.debug(
                     "[PIPELINE:UMA][TRANSFORM] req=%s stage=ai_chat_messages done applied=%s skipped=%s reasons=%s",
                     req_id,
                     transform_diag.get("applied_count", 0),
@@ -5764,7 +5764,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                     pre_contact_len,
                 )
             transform_diag = {}
-            logger.info(
+            logger.debug(
                 "[PIPELINE:UMA][TRANSFORM] req=%s stage=ai_chat_messages_jsonl start rows=%s",
                 req_id,
                 len(messages),
@@ -5777,7 +5777,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                 diagnostics=transform_diag,
                 progress_hook=_uma_transform_progress_hook(req_id, "ai_chat_messages_jsonl"),
             )
-            logger.info(
+            logger.debug(
                 "[PIPELINE:UMA][TRANSFORM] req=%s stage=ai_chat_messages_jsonl done applied=%s skipped=%s reasons=%s",
                 req_id,
                 transform_diag.get("applied_count", 0),
@@ -5951,7 +5951,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
             rows = all_rows[:limit]
             try:
                 transform_diag = {}
-                logger.info(
+                logger.debug(
                     "[PIPELINE:UMA][TRANSFORM] req=%s stage=rows:%s start rows=%s",
                     req_id,
                     table_name,
@@ -5965,7 +5965,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                     diagnostics=transform_diag,
                     progress_hook=_uma_transform_progress_hook(req_id, f"rows:{table_name}"),
                 )
-                logger.info(
+                logger.debug(
                     "[PIPELINE:UMA][TRANSFORM] req=%s stage=rows:%s done applied=%s skipped=%s reasons=%s",
                     req_id,
                     table_name,
@@ -6840,7 +6840,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
         sync_options = payload.get("sync_options")
         if source_id == "imessage" and not sync_options:
             sync_options = {"mode": "3m"}
-        logger.info(
+        logger.debug(
             "[PIPELINE:SYNC] Engine received source_sync request: source_id=%s dataset_id=%s sync_options=%s (req_id=%s)",
             source_id,
             dataset_id[:24] + "..." if dataset_id and len(dataset_id) > 24 else dataset_id,
@@ -6870,7 +6870,7 @@ async def handle_control_plane_request(message: Dict[str, Any]) -> Optional[Dict
                     error,
                     exc_info=False,
                 )
-            logger.info("[PIPELINE:SYNC] source_sync completed: source_id=%s status=%s", source_id, status)
+            logger.debug("[PIPELINE:SYNC] source_sync completed: source_id=%s status=%s", source_id, status)
             if conn and status == "ok":
                 update_sync_result(conn, dataset_id, source_id, success=True, last_sync_at=datetime.now(timezone.utc).isoformat())
             elif conn and status == "error":

{topos_node-1.0.0 → topos_node-1.0.2}/topos/core/logging.py

@@ -1,5 +1,7 @@
 import json
 import logging
+import os
+import re
 import sys
 from datetime import datetime
 from time import time
@@ -9,6 +11,26 @@ from ..runtime_update import is_update_available
 
 _LOG_FORMAT: str | None = None
 
+# Before huggingface_hub/transformers import, suppress download/load tqdm bars.
+os.environ.setdefault("HF_HUB_DISABLE_PROGRESS_BARS", "1")
+
+
+def suppress_ml_progress_bars() -> None:
+    """Hide Hugging Face / transformers tqdm progress bars; keep app INFO load logs."""
+    os.environ.setdefault("HF_HUB_DISABLE_PROGRESS_BARS", "1")
+    try:
+        from huggingface_hub.utils import disable_progress_bars
+
+        disable_progress_bars()
+    except Exception:
+        pass
+    try:
+        from transformers.utils.logging import disable_progress_bar
+
+        disable_progress_bar()
+    except Exception:
+        pass
+
 # Route uvicorn loggers through the root handler configured by configure_logging().
 UVICORN_LOG_CONFIG: dict[str, object] = {
     "version": 1,
@@ -37,6 +59,7 @@ def align_uvicorn_loggers() -> None:
 
 def configure_logging() -> None:
     """Configure logging to stdout based on environment/log settings."""
+    suppress_ml_progress_bars()
     handler = logging.StreamHandler(sys.stdout)
 
     def resolve_format() -> str:
@@ -145,12 +168,42 @@ class ColorFormatter(logging.Formatter):
     
     # Message text: Same blue as DEBUG
     _MESSAGE_COLOR = "\x1b[38;5;26m"  # Light blue (same as DEBUG)
+
+    # source / source_id values: bold magenta — easy to scan in pipeline logs
+    _SOURCE_IDENTITY_VALUE_COLOR = "\x1b[1;38;5;213m"  # Bold bright magenta
     
     # Logger name (module path): Same teal/cyan as class names
     _LOGGER_NAME_COLOR = "\x1b[38;5;75m"  # Teal/cyan blue-green
     
     _ARROW_COLOR = "\x1b[38;5;244m"  # Gray for arrows
 
+    _SOURCE_KV_PATTERN = re.compile(
+        r"(?<![?&/])"  # skip URL query/path fragments like ...&source=editors
+        r"\b(?P<key>source_id|source)="
+        r"(?:"
+        r"(?P<q>['\"])(?P<qval>[^'\"]+)(?P=q)|"
+        r"(?P<uval>[^\s,;)\]&]+)"
+        r")"
+    )
+
+    _SOURCE_VALUE_SKIP = frozenset({"%s", "%d", "%i", "%f", "?"})
+
+    def _highlight_source_identities(self, message: str) -> str:
+        """Bold magenta on source / source_id values (keys stay default message color)."""
+
+        def repl(match: re.Match[str]) -> str:
+            value = match.group("qval") or match.group("uval") or ""
+            if value in self._SOURCE_VALUE_SKIP or value.startswith("excluded."):
+                return match.group(0)
+            key = match.group("key")
+            quote = match.group("q") or ""
+            colored_value = (
+                f"{self._SOURCE_IDENTITY_VALUE_COLOR}{value}{self._RESET}{self._MESSAGE_COLOR}"
+            )
+            return f"{key}={quote}{colored_value}{quote}"
+
+        return self._SOURCE_KV_PATTERN.sub(repl, message)
+
     def format(self, record: logging.LogRecord) -> str:  # noqa: D401
         # Format timestamp: YYYY-MM-DD HH:MM:SS.ms (green)
         timestamp_str = datetime.fromtimestamp(record.created).strftime("%Y-%m-%d %H:%M:%S.%f")[:-3]
@@ -171,8 +224,7 @@ class ColorFormatter(logging.Formatter):
         
         # Get message and extract pipeline tag if present
         message = record.getMessage()
-        import re
-        
+
         # Extract pipeline stage tag from message if it exists
         pipeline_tag_pattern = r'\[PIPELINE:[A-Z_]+\]'
         pipeline_tag_match = re.search(pipeline_tag_pattern, message)
@@ -204,6 +256,8 @@ class ColorFormatter(logging.Formatter):
             f"{self._ARROW_COLOR}\\1{self._RESET}{self._MESSAGE_COLOR}",
             message_colored
         )
+
+        message_colored = self._highlight_source_identities(message_colored)
         
         # Format: TIMESTAMP | LEVEL | [PIPELINE:TAG] | LOGGER: MESSAGE
         separator = f"{self._SEPARATOR_COLOR}|{self._RESET}"

{topos_node-1.0.0 → topos_node-1.0.2}/topos/disclosure/privacy_layer.py

@@ -184,11 +184,11 @@ async def run_privacy_disclosure_layer(
     apply_canonical_nsfw_v1_up(conn)
 
     if not getattr(settings, "platform_privacy_via_engine", True):
-        logger.info("[PIPELINE:PRIVACY] skipped: platform_privacy_via_engine=false")
+        logger.debug("[PIPELINE:PRIVACY] skipped: platform_privacy_via_engine=false")
         return {"records_updated": 0, "skipped": True}
 
     if nsfw_only and not getattr(settings, "nsfw_classifier_enabled", True):
-        logger.info("[PIPELINE:PRIVACY] nsfw-only backfill skipped: nsfw_classifier_enabled=false")
+        logger.debug("[PIPELINE:PRIVACY] nsfw-only backfill skipped: nsfw_classifier_enabled=false")
         return {"records_updated": 0, "nsfw_tagged": 0, "skipped": True}
 
     if not canonical_messages:
@@ -353,7 +353,7 @@ async def run_privacy_disclosure_layer(
         conn.commit()
 
     duration_ms = int((time.perf_counter() - started) * 1000)
-    logger.info(
+    logger.debug(
         "[PIPELINE:PRIVACY] platform_privacy_layer disclosure_updated=%d nsfw_tagged=%d "
         "failed_batches=%d nsfw_failed_batches=%d duration_ms=%d version=%s",
         updated,

{topos_node-1.0.0 → topos_node-1.0.2}/topos/engine/backends/huggingface.py

@@ -278,7 +278,12 @@ class HuggingFaceAdapter:
         vectors: List[List[float]] = []
         for i in range(0, len(texts), batch_size):
             batch = texts[i : i + batch_size]
-            encoded = embedder.encode(batch, convert_to_numpy=True, normalize_embeddings=normalize)
+            encoded = embedder.encode(
+                batch,
+                convert_to_numpy=True,
+                normalize_embeddings=normalize,
+                show_progress_bar=False,
+            )
             for row in encoded:
                 vectors.append([float(x) for x in row.tolist()])
         dims = len(vectors[0]) if vectors else 0

{topos_node-1.0.0 → topos_node-1.0.2}/topos/enrichment/jobs/canonical/topic_clusters_job.py

@@ -41,7 +41,7 @@ class TopicClusterJob(BaseEnrichmentJob):
         # does not wipe cross-source memory_topic_map rollups.
         scope_ids = list(_resolved_topic_cluster_source_ids())
         if source_ids:
-            logger.info(
+            logger.debug(
                 "[PIPELINE:TOPIC_CLUSTERS] batch sources=%s; clustering scope=%s",
                 sorted(source_ids),
                 scope_ids,
@@ -57,7 +57,7 @@ class TopicClusterJob(BaseEnrichmentJob):
             min_records=3,
         )
         if result.get("status") == "skipped":
-            logger.info("[PIPELINE:TOPIC_CLUSTERS] skipped: %s", result.get("reason"))
+            logger.debug("[PIPELINE:TOPIC_CLUSTERS] skipped: %s", result.get("reason"))
             if progress_callback:
                 progress_callback(1, 1)
             return []

{topos_node-1.0.0 → topos_node-1.0.2}/topos/enrichment/orchestrator.py

@@ -58,7 +58,7 @@ class EnrichmentOrchestrator(BaseObject):
         
         total_messages = len(canonical_messages)
         total_jobs = len(jobs_to_run)
-        logger.info(
+        logger.debug(
             "[PIPELINE:ENRICHMENT] %s: Starting enrichment: %d messages, %d jobs to run",
             self,
             total_messages,
@@ -75,7 +75,7 @@ class EnrichmentOrchestrator(BaseObject):
                 continue
             try:
                 job_name = job.get_job_name()
-                logger.info(
+                logger.debug(
                     "[PIPELINE:ENRICHMENT] %s: Running job %d/%d: %s (%d messages, %.1f%% of jobs complete)",
                     self,
                     job_idx,
@@ -115,7 +115,7 @@ class EnrichmentOrchestrator(BaseObject):
                         records, derived_table
                     )
                     results["records_created"][derived_table] = records_written
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:ENRICHMENT] %s → %s: %d records written to %s (job %d/%d, %.1f%% complete)",
                         self,
                         job,
@@ -132,7 +132,7 @@ class EnrichmentOrchestrator(BaseObject):
                     results["records_created"]["canonical_disclosure"] = (
                         int(results["records_created"].get("canonical_disclosure") or 0) + updated
                     )
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:ENRICHMENT] %s → %s: %d canonical disclosure rows updated (job %d/%d)",
                         self,
                         job,
@@ -143,7 +143,7 @@ class EnrichmentOrchestrator(BaseObject):
                 else:
                     table_name = derived_table or job.get_job_name()
                     results["records_created"][table_name] = 0
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:ENRICHMENT] %s → %s: completed with 0 records (job %d/%d, %.1f%% complete)",
                         self,
                         job,
@@ -181,7 +181,7 @@ class EnrichmentOrchestrator(BaseObject):
                     # Mark job as 100% complete (even if failed, we've moved past it)
                     progress_callback(messages_processed_so_far, total_messages, job.get_job_name(), job_progress_percent, 100.0)
         
-        logger.info(
+        logger.debug(
             "[PIPELINE:ENRICHMENT] %s: Enrichment complete: %d jobs run, %d total records created",
             self,
             results["jobs_run"],
@@ -275,7 +275,7 @@ class SignalDerivationOrchestrator(EnrichmentOrchestrator):
         }
         jobs_to_run = [self._signal_jobs[name] for name in resolved if name in self._signal_jobs]
         total_jobs = len(jobs_to_run)
-        logger.info(
+        logger.debug(
             "[PIPELINE:SIGNAL_DERIVE] source_id=%s batch_id=%s jobs=%s messages=%d",
             source_id,
             sync_batch_id,
@@ -343,7 +343,7 @@ class SignalDerivationOrchestrator(EnrichmentOrchestrator):
         except Exception as exc:
             logger.debug("[PIPELINE:SIGNAL_DERIVE] dimension profile update skipped: %s", exc)
 
-        logger.info(
+        logger.debug(
             "[PIPELINE:SIGNAL_DERIVE] complete source_id=%s batch_id=%s jobs_run=%d deferred=%s",
             source_id,
             sync_batch_id,

{topos_node-1.0.0 → topos_node-1.0.2}/topos/ingestion/canonical_pipeline.py

@@ -683,7 +683,7 @@ async def run_post_canonical_pipeline(
             logger.error("[PIPELINE:ENRICHMENT] post-canonical failed: %s", exc, exc_info=True)
             outcome["canonical_enrichment"] = {"errors": [str(exc)]}
     elif canonical_jobs and enrichment_trigger == "manual":
-        logger.info(
+        logger.debug(
             "[PIPELINE:ENRICHMENT] Skipping canonical enrichment (manual trigger): source_id=%s records=%d",
             source_id,
             len(canonical_records),

{topos_node-1.0.0 → topos_node-1.0.2}/topos/ingestion/ingest_helpers.py

@@ -86,7 +86,7 @@ async def ingest_file_payload(
         for source in REGISTRY.values():
             if source.schema_id == schema_id and source.source_type == "file":
                 source_id = source.source_id
-                logger.debug(
+                logger.info(
                     "[PIPELINE:RAW] Found source_id=%s for schema_id=%s (file type)",
                     source_id,
                     schema_id,
@@ -101,7 +101,7 @@ async def ingest_file_payload(
     file_store = RawFileStore()
     trigger = FileTrigger(file_store=file_store)
     job_id = job_id or str(uuid.uuid4())
-    logger.debug(
+    logger.info(
         "[PIPELINE:RAW] Creating file ingestion job: job_id=%s, dataset_id=%s, schema_id=%s, source_id=%s, file_path=%s, file_size=%s",
         job_id,
         dataset_id,
@@ -123,7 +123,7 @@ async def ingest_file_payload(
         progress_api_url=progress_api_url,
         progress_api_key=progress_api_key,
     )
-    logger.debug(
+    logger.info(
         "[PIPELINE:RAW] %s: File ingestion complete: job_id=%s, records_processed=%s, errors=%s",
         manager,
         job_id,
@@ -201,7 +201,7 @@ async def ingest_ui_payload(
         "content": content,
         "created_at": created_at,
     }
-    logger.debug(
+    logger.info(
         "[PIPELINE:RAW] Appending UI message to raw store: job_id=%s, dataset_id=%s, message_id=%s, content_preview=%s",
         job_id,
         dataset_id,
@@ -216,10 +216,10 @@ async def ingest_ui_payload(
         schema_id,
         file_store.get_file_path(dataset_id, schema_id).as_posix(),
     )
-    logger.debug("[PIPELINE:RAW] Starting ingestion job: job_id=%s, dataset_id=%s, schema_id=%s", job_id, dataset_id, schema_id)
+    logger.info("[PIPELINE:RAW] Starting ingestion job: job_id=%s, dataset_id=%s, schema_id=%s", job_id, dataset_id, schema_id)
     manager = IngestionManager(file_store=file_store)
     result = await manager.process_job(job)
-    logger.debug(
+    logger.info(
         "[PIPELINE:RAW] %s: UI ingestion complete: job_id=%s, records_processed=%s, errors=%s",
         manager,
         job_id,
@@ -389,7 +389,7 @@ async def _ingest_ui_payload_direct(
             payload=raw_payload,
             source_type=raw_source_type,
         )
-        logger.debug(
+        logger.info(
             "[PIPELINE:RAW] Stored raw payload: source=%s, record_id=%s",
             source_id,
             record_id[:8] if record_id else None,

{topos_node-1.0.0 → topos_node-1.0.2}/topos/ingestion/manager.py

@@ -587,7 +587,7 @@ async def _try_install_runtime_source_definition_from_control_plane(
             install_source_definition(row)
             installed = REGISTRY.get(row_source_id)
             if installed:
-                logger.info(
+                logger.debug(
                     "[PIPELINE:MANAGER] Installed runtime source definition from control-plane: source_id=%s schema_id=%s",
                     row_source_id,
                     row_schema_id,
@@ -690,7 +690,7 @@ class IngestionManager(BaseObject):
         if source_id:
             source_def = REGISTRY.get(source_id)
             if source_def:
-                logger.info(
+                logger.debug(
                     "[PIPELINE:MANAGER] %s: Using source from source_id=%s: %s (enrichment_trigger=%s)",
                     self,
                     source_id,
@@ -745,7 +745,7 @@ class IngestionManager(BaseObject):
                         # Keep first match as fallback
                         source_def = source
             if source_def:
-                logger.info(
+                logger.debug(
                     "[PIPELINE:MANAGER] %s: Found source by schema_id: source_id=%s, source_type=%s, enrichment_trigger=%s",
                     self,
                     source_def.source_id,
@@ -1009,7 +1009,7 @@ class IngestionManager(BaseObject):
                     and enrichment_trigger == "automatic"
                     and tables_manager
                 ):
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:ENRICHMENT] %s: Enrichment trigger check: source_id=%s, enrichment_trigger=%s, canonical_messages=%d, jobs=%s",
                         self,
                         source_def.source_id,
@@ -1031,7 +1031,7 @@ class IngestionManager(BaseObject):
                             len(canonical_messages),
                         )
                 elif source_def.canonical_enrichment_jobs and enrichment_trigger == "manual":
-                    logger.info(
+                    logger.debug(
                         "[PIPELINE:ENRICHMENT] %s: Skipping enrichment (manual trigger): %d canonical messages will be enriched later via POST /v1/enrichment/process",
                         self,
                         len(canonical_messages),