Auto-adds internal_analytics. prefix to table names Skill

ClickHouse Internal Analytics = Palantir Shard Access

Core Principle: Query internal_analytics via internal.query(). Use customer.query() for client data.

Architecture:

graph TD
    CH[ClickHouse] --> P[Palantir]
    CH --> M[Montana]

    P --> IA[internal_analytics]
    M --> DTS[DTS clients via MCP]
    M --> RTBM[RTBM-only clients direct]

Routing:

graph LR
    Q[Query] --> T{Data type?}
    T -->|internal_analytics| I[internal.query]
    T -->|client data| C[customer.query]
    C --> R{Client type?}
    R -->|DTS| MCP[MCP]
    R -->|RTBM-only| Direct[Montana direct]

Primary source: data_sources/clickhouse/ Session: 2026-01-01 - v3.1.0 RTBM-only support

Query Tools

¶1 CLI:

ch internal "SELECT * FROM biz_active_customers LIMIT 5"
ch internal "SELECT name FROM system.tables WHERE database = 'internal_analytics'"

¶2 Python:

from data_sources.clickhouse import internal

df = internal.query("SELECT * FROM biz_active_customers LIMIT 5")
df = internal.query("SELECT name FROM system.tables WHERE database = 'internal_analytics'")

# Auto-adds internal_analytics. prefix to table names
df = internal.query("SELECT * FROM dim_gong_calls LIMIT 5")

Key Tables

| Table | Purpose | Key Columns | |-------|---------|-------------| | biz_active_customers | Customers | dts_agency_id, dts_agency_title, company_domain_id | | dim_jira_issues | Jira tickets | jira_ticket_id, jira_domain_id, jira_ticket_summary | | dim_gong_calls | Gong calls | gong_call_url, gong_call_title, gong_call_started | | dim_accounts | Accounts | account_id, account_name, mrr |

Common Queries

¶1 Find customer:

ch find hp                    # Quick CLI search

from data_sources.clickhouse import internal

df = internal.query("""
    SELECT dts_agency_id, dts_agency_title, company_domain_id
    FROM biz_active_customers
    WHERE lower(dts_agency_name) LIKE '%customer%'
""")

¶2 Customer Jira tickets:

df = internal.query("""
    SELECT jira_ticket_id, jira_ticket_summary, jira_ticket_status
    FROM dim_jira_issues
    WHERE jira_domain_id = 'customer.org'
""")

¶3 Schema discovery:

# List tables
df = internal.query("SELECT name FROM system.tables WHERE database = 'internal_analytics' AND name LIKE '%jira%'")

# Describe table
df = internal.query("DESCRIBE TABLE internal_analytics.dim_jira_issues")

Ground Truth

CLI: ch internal "SQL"
Python: from data_sources.clickhouse import internal; df = internal.query("SQL")
Database: internal_analytics on Palantir shard

Customer Data (NOT here)

from data_sources.clickhouse import customer

# DTS client (uses MCP)
df = customer.query("SELECT * FROM facebook", agency="hp.com")

# RTBM-only client (uses Montana direct)
df = customer.query("SELECT * FROM keywords_21_botify", agency="t-mobile.com")

# Find agency info
info = customer.resolve("hp")  # → {'agency_id': 19277, 'is_rtbm_only': False, ...}

Agent Skills: Auto-adds internal_analytics. prefix to table names