Merge pull request #34 from hackclub/revert-4ed00a2673b4aaaa4554aab929ac

Author: maxwofford

warehouse-mcp/db.py

@@ -9,10 +9,6 @@
 from typing import List, Dict, Any, Optional
 from psycopg.rows import dict_row
 from psycopg_pool import AsyncConnectionPool
-from warehouse_docs import (
-    UNIFIED_ANALYTICS_EVENTS_GUIDANCE,
-    UNIFIED_ANALYTICS_EVENTS_SQL_EXAMPLES,
-)
 
 
 # SQL statements that are never allowed, even with read-only connection
@@ -431,13 +427,7 @@ async def describe_schema(
             return f"Schema '{schema_name}' not found or has no tables."
 
         output_parts = [f"# Schema: {schema_name}\n"]
-        if schema_name == "public_unified_analytics":
-            output_parts.append(
-                "\n## Signup Analytics Guidance\n"
-                f"{UNIFIED_ANALYTICS_EVENTS_GUIDANCE}\n\n"
-                f"{UNIFIED_ANALYTICS_EVENTS_SQL_EXAMPLES}\n"
-            )
-        current_size = sum(len(part.encode('utf-8')) for part in output_parts)
+        current_size = len(output_parts[0])
         budget_exceeded = False
 
         for table_row in tables:

warehouse-mcp/server.py

@@ -29,7 +29,6 @@
 import psycopg
 from db import get_database, init_database, close_database, SQLValidationError
 from cache import get_cache
-from warehouse_docs import UNIFIED_ANALYTICS_EVENTS_GUIDANCE
 
 # Load environment variables
 load_dotenv()
@@ -245,8 +244,7 @@ async def list_tools() -> list[Tool]:
             description=(
                 "Execute a read-only SQL query against the data warehouse. "
                 "Returns a query_id for fetching more rows, plus a preview of results. "
-                "Use LIMIT in your SQL for large tables. Max 10,000 rows cached per query.\n\n"
-                f"Analytics guidance: {UNIFIED_ANALYTICS_EVENTS_GUIDANCE}"
+                "Use LIMIT in your SQL for large tables. Max 10,000 rows cached per query."
             ),
             inputSchema={
                 "type": "object",
@@ -323,11 +321,8 @@ async def list_tools() -> list[Tool]:
                 "List all available schemas in the data warehouse.\n\n"
                 "Key schemas:\n"
                 "- airtable_unified_ysws_projects_db: Shipped projects and NPS on programs\n"
-                "- public_unified_analytics: User histories and events on users. "
-                "Use public_unified_analytics.events for signups and signup sources; "
-                "the first event per normalized email is the signup.\n"
-                "- loops.audience: User profile fields (geocoded country, etc.); "
-                "do not use loops.audience.source for signup/source attribution\n"
+                "- public_unified_analytics: User histories and events on users\n"
+                "- loops.audience: User profile fields (geocoded country, etc.)\n"
                 "- public_hackatime_analytics: Coding activity (preferred over hackatime schema)"
             ),
             inputSchema={
@@ -341,10 +336,7 @@ async def list_tools() -> list[Tool]:
             description=(
                 "Get detailed documentation for a schema including tables, columns, "
                 "and sample data. Limited to 1000 columns per table. "
-                "Use list_columns for tables with more columns. "
-                "For signup counts and signup sources, describe public_unified_analytics "
-                "and query public_unified_analytics.events; the first event per normalized "
-                "email is the signup."
+                "Use list_columns for tables with more columns."
             ),
             inputSchema={
                 "type": "object",