Snowflake Development
Snowflake SQL, data pipelines, Cortex AI, and Snowpark Python development. Covers the colon-prefix rule, semi-structured data, MERGE upserts, Dynamic Tables, Streams+Tasks, Cortex AI functions, agent specs, performance tuning, and security hardening.
Originally contributed by James Cha-Earley — enhanced and integrated by the claude-skills team.
Quick Start
# Generate a MERGE upsert template
python scripts/snowflake_query_helper.py merge --target customers --source staging_customers --key customer_id --columns name,email,updated_at
# Generate a Dynamic Table template
python scripts/snowflake_query_helper.py dynamic-table --name cleaned_events --warehouse transform_wh --lag "5 minutes"
# Generate RBAC grant statements
python scripts/snowflake_query_helper.py grant --role analyst_role --database analytics --schemas public,staging --privileges SELECT,USAGE
SQL Best Practices
Naming and Style
- Use
snake_case for all identifiers. Avoid double-quoted identifiers -- they force case-sensitive names that require constant quoting.
- Use CTEs (
WITH clauses) over nested subqueries.
- Use
CREATE OR REPLACE for idempotent DDL.
- Use explicit column lists -- never
SELECT * in production. Snowflake's columnar storage scans only referenced columns, so explicit lists reduce I/O.
Stored Procedures -- Colon Prefix Rule
In SQL stored procedures (BEGIN...END blocks), variables and parameters must use the colon : prefix inside SQL statements. Without it, Snowflake treats them as column identifiers and raises "invalid identifier" errors.
-- WRONG: missing colon prefix
SELECT name INTO result FROM users WHERE id = p_id;
-- CORRECT: colon prefix on both variable and parameter
SELECT name INTO :result FROM users WHERE id = :p_id;
This applies to DECLARE variables, LET variables, and procedure parameters when used inside SELECT, INSERT, UPDATE, DELETE, or MERGE.
Semi-Structured Data
- VARIANT, OBJECT, ARRAY for JSON/Avro/Parquet/ORC.
- Access nested fields:
src:customer.name::STRING. Always cast with ::TYPE.
- VARIANT null vs SQL NULL: JSON
null is stored as the string "null". Use STRIP_NULL_VALUE = TRUE on load.
- Flatten arrays:
SELECT f.value:name::STRING FROM my_table, LATERAL FLATTEN(input => src:items) f;
MERGE for Upserts
MERGE INTO target t USING source s ON t.id = s.id
WHEN MATCHED THEN UPDATE SET t.name = s.name, t.updated_at = CURRENT_TIMESTAMP()
WHEN NOT MATCHED THEN INSERT (id, name, updated_at) VALUES (s.id, s.name, CURRENT_TIMESTAMP());
See references/snowflake_sql_and_pipelines.md for deeper SQL patterns and anti-patterns.
Data Pipelines
Choosing Your Approach
| Approach |
When to Use |
| Dynamic Tables |
Declarative transformations. Default choice. Define the query, Snowflake handles refresh. |
| Streams + Tasks |
Imperative CDC. Use for procedural logic, stored procedure calls, complex branching. |
| Snowpipe |
Continuous file loading from cloud storage (S3, GCS, Azure). |
Dynamic Tables
CREATE OR REPLACE DYNAMIC TABLE cleaned_events
TARGET_LAG = '5 minutes'
WAREHOUSE = transform_wh
AS
SELECT event_id, event_type, user_id, event_timestamp
FROM raw_events
WHERE event_type IS NOT NULL;
Key rules:
- Set
TARGET_LAG progressively: tighter at the top of the DAG, looser downstream.
- Incremental DTs cannot depend on Full-refresh DTs.
SELECT * breaks on upstream schema changes -- use explicit column lists.
- Views cannot sit between two Dynamic Tables in the DAG.
Streams and Tasks
CREATE OR REPLACE STREAM raw_stream ON TABLE raw_events;
CREATE OR REPLACE TASK process_events
WAREHOUSE = transform_wh
SCHEDULE = 'USING CRON 0 */1 * * * America/Los_Angeles'
WHEN SYSTEM$STREAM_HAS_DATA('raw_stream')
AS INSERT INTO cleaned_events SELECT ... FROM raw_stream;
-- Tasks start SUSPENDED. You MUST resume them.
ALTER TASK process_events RESUME;
See references/snowflake_sql_and_pipelines.md for DT debugging queries and Snowpipe patterns.
Cortex AI
Function Reference
| Function |
Purpose |
AI_COMPLETE |
LLM completion (text, images, documents) |
AI_CLASSIFY |
Classify text into categories (up to 500 labels) |
AI_FILTER |
Boolean filter on text or images |
AI_EXTRACT |
Structured extraction from text/images/documents |
AI_SENTIMENT |
Sentiment score (-1 to 1) |
AI_PARSE_DOCUMENT |
OCR or layout extraction from documents |
AI_REDACT |
PII removal from text |
Deprecated names (do NOT use): COMPLETE, CLASSIFY_TEXT, EXTRACT_ANSWER, PARSE_DOCUMENT, SUMMARIZE, TRANSLATE, SENTIMENT, EMBED_TEXT_768.
TO_FILE -- Common Pitfall
Stage path and filename are separate arguments:
-- WRONG: single combined argument
TO_FILE('@stage/file.pdf')
-- CORRECT: two arguments
TO_FILE('@db.schema.mystage', 'invoice.pdf')
Cortex Agents
Agent specs use a JSON structure with top-level keys: models, instructions, tools, tool_resources.
- Use
$spec$ delimiter (not $$).
models must be an object, not an array.
tool_resources is a separate top-level key, not nested inside tools.
- Tool descriptions are the single biggest factor in agent quality.
See references/cortex_ai_and_agents.md for full agent spec examples and Cortex Search patterns.
Snowpark Python
from snowflake.snowpark import Session
import os
session = Session.builder.configs({
"account": os.environ["SNOWFLAKE_ACCOUNT"],
"user": os.environ["SNOWFLAKE_USER"],
"password": os.environ["SNOWFLAKE_PASSWORD"],
"role": "my_role", "warehouse": "my_wh",
"database": "my_db", "schema": "my_schema"
}).create()
- Never hardcode credentials. Use environment variables or key pair auth.
- DataFrames are lazy -- executed on
collect() / show().
- Do NOT call
collect() on large DataFrames. Process server-side with DataFrame operations.
- Use vectorized UDFs (10-100x faster) for batch and ML workloads.
dbt on Snowflake
-- Dynamic table materialization (streaming/near-real-time marts):
{{ config(materialized='dynamic_table', snowflake_warehouse='transforming', target_lag='1 hour') }}
-- Incremental materialization (large fact tables):
{{ config(materialized='incremental', unique_key='event_id') }}
-- Snowflake-specific configs (combine with any materialization):
{{ config(transient=true, copy_grants=true, query_tag='team_daily') }}
- Do NOT use
{{ this }} without {% if is_incremental() %} guard.
- Use
dynamic_table materialization for streaming or near-real-time marts.
Performance
- Cluster keys: Only for multi-TB tables. Apply on WHERE / JOIN / GROUP BY columns.
- Search Optimization:
ALTER TABLE t ADD SEARCH OPTIMIZATION ON EQUALITY(col);
- Warehouse sizing: Start X-Small, scale up. Set
AUTO_SUSPEND = 60, AUTO_RESUME = TRUE.
- Separate warehouses per workload (load, transform, query).
Security
- Follow least-privilege RBAC. Use database roles for object-level grants.
- Audit ACCOUNTADMIN regularly:
SHOW GRANTS OF ROLE ACCOUNTADMIN;
- Use network policies for IP allowlisting.
- Use masking policies for PII columns and row access policies for multi-tenant isolation.
Proactive Triggers
Surface these issues without being asked when you notice them in context:
- Missing colon prefix in SQL stored procedures -- flag immediately, this causes "invalid identifier" at runtime.
SELECT * in Dynamic Tables -- flag as a schema-change time bomb.
- Deprecated Cortex function names (
CLASSIFY_TEXT, SUMMARIZE, etc.) -- suggest the current AI_* equivalents.
- Task not resumed after creation -- remind that tasks start SUSPENDED.
- Hardcoded credentials in Snowpark code -- flag as a security risk.
Common Errors
| Error |
Cause |
Fix |
| "Object does not exist" |
Wrong database/schema context or missing grants |
Fully qualify names (db.schema.table), check grants |
| "Invalid identifier" in procedure |
Missing colon prefix on variable |
Use :variable_name inside SQL statements |
| "Numeric value not recognized" |
VARIANT field not cast |
Cast explicitly: src:field::NUMBER(10,2) |
| Task not running |
Forgot to resume after creation |
ALTER TASK task_name RESUME; |
| DT refresh failing |
Schema change upstream or tracking disabled |
Use explicit columns, verify change tracking |
| TO_FILE error |
Combined path as single argument |
Split into two args: TO_FILE('@stage', 'file.pdf') |
Practical Workflows
Workflow 1: Build a Reporting Pipeline (30 min)
- Stage raw data: Create external stage pointing to S3/GCS/Azure, set up Snowpipe for auto-ingest
- Clean with Dynamic Table: Create DT with
TARGET_LAG = '5 minutes' that filters nulls, casts types, deduplicates
- Aggregate with downstream DT: Second DT that joins cleaned data with dimension tables, computes metrics
- Expose via Secure View: Create
SECURE VIEW for the BI tool / API layer
- Grant access: Use
snowflake_query_helper.py grant to generate RBAC statements
Workflow 2: Add AI Classification to Existing Data
- Identify the column: Find the text column to classify (e.g., support tickets, reviews)
- Test with AI_CLASSIFY:
SELECT AI_CLASSIFY(text_col, ['bug', 'feature', 'question']) FROM table LIMIT 10;
- Create enrichment DT: Dynamic Table that runs
AI_CLASSIFY on new rows automatically
- Monitor costs: Cortex AI is billed per token — sample before running on full tables
Workflow 3: Debug a Failing Pipeline
- Check task history:
SELECT * FROM TABLE(INFORMATION_SCHEMA.TASK_HISTORY()) WHERE STATE = 'FAILED' ORDER BY SCHEDULED_TIME DESC;
- Check DT refresh:
SELECT * FROM TABLE(INFORMATION_SCHEMA.DYNAMIC_TABLE_REFRESH_HISTORY('my_dt')) ORDER BY REFRESH_END_TIME DESC;
- Check stream staleness:
SHOW STREAMS; -- check stale_after column
- Consult troubleshooting reference: See
references/troubleshooting.md for error-specific fixes
Anti-Patterns
| Anti-Pattern |
Why It Fails |
Better Approach |
SELECT * in Dynamic Tables |
Schema changes upstream break the DT silently |
Use explicit column lists |
| Missing colon prefix in procedures |
"Invalid identifier" runtime error |
Always use :variable_name in SQL blocks |
| Single warehouse for all workloads |
Contention between load, transform, and query |
Separate warehouses per workload type |
| Hardcoded credentials in Snowpark |
Security risk, breaks in CI/CD |
Use os.environ[] or key pair auth |
collect() on large DataFrames |
Pulls entire result set to client memory |
Process server-side with DataFrame operations |
| Nested subqueries instead of CTEs |
Unreadable, hard to debug, Snowflake optimizes CTEs better |
Use WITH clauses |
| Using deprecated Cortex functions |
CLASSIFY_TEXT, SUMMARIZE etc. will be removed |
Use AI_CLASSIFY, AI_COMPLETE etc. |
Tasks without WHEN SYSTEM$STREAM_HAS_DATA |
Task runs on schedule even with no new data, wasting credits |
Add the WHEN clause for stream-driven tasks |
| Double-quoted identifiers |
Forces case-sensitive names across all queries |
Use snake_case unquoted identifiers |
Cross-References
| Skill |
Relationship |
engineering/sql-database-assistant |
General SQL patterns — use for non-Snowflake databases |
engineering/database-designer |
Schema design — use for data modeling before Snowflake implementation |
engineering-team/senior-data-engineer |
Broader data engineering — pipelines, Spark, Airflow, data quality |
engineering-team/senior-data-scientist |
Analytics and ML — use alongside Snowpark for feature engineering |
engineering-team/senior-devops |
CI/CD for Snowflake deployments (Terraform, GitHub Actions) |
Reference Documentation
| Document |
Contents |
references/snowflake_sql_and_pipelines.md |
SQL patterns, MERGE templates, Dynamic Table debugging, Snowpipe, anti-patterns |
references/cortex_ai_and_agents.md |
Cortex AI functions, agent spec structure, Cortex Search, Snowpark |
references/troubleshooting.md |
Error reference, debugging queries, common fixes |
1---2name: snowflake-development3description: "Use when writing Snowflake SQL, building data pipelines with Dynamic Tables or Streams/Tasks, using Cortex AI functions, creating Cortex Agents, writing Snowpark Python, configuring dbt for Snowflake, or troubleshooting Snowflake errors."4metadata:5 source: https://skillmd.com/skills/bestagentkits/snowflake-development6---78# Snowflake Development910Snowflake SQL, data pipelines, Cortex AI, and Snowpark Python development. Covers the colon-prefix rule, semi-structured data, MERGE upserts, Dynamic Tables, Streams+Tasks, Cortex AI functions, agent specs, performance tuning, and security hardening.1112> Originally contributed by [James Cha-Earley](https://github.com/jamescha-earley) — enhanced and integrated by the claude-skills team.1314## Quick Start1516```bash17# Generate a MERGE upsert template18python scripts/snowflake_query_helper.py merge --target customers --source staging_customers --key customer_id --columns name,email,updated_at1920# Generate a Dynamic Table template21python scripts/snowflake_query_helper.py dynamic-table --name cleaned_events --warehouse transform_wh --lag "5 minutes"2223# Generate RBAC grant statements24python scripts/snowflake_query_helper.py grant --role analyst_role --database analytics --schemas public,staging --privileges SELECT,USAGE25```2627---2829## SQL Best Practices3031### Naming and Style3233- Use `snake_case` for all identifiers. Avoid double-quoted identifiers -- they force case-sensitive names that require constant quoting.34- Use CTEs (`WITH` clauses) over nested subqueries.35- Use `CREATE OR REPLACE` for idempotent DDL.36- Use explicit column lists -- never `SELECT *` in production. Snowflake's columnar storage scans only referenced columns, so explicit lists reduce I/O.3738### Stored Procedures -- Colon Prefix Rule3940In SQL stored procedures (BEGIN...END blocks), variables and parameters **must** use the colon `:` prefix inside SQL statements. Without it, Snowflake treats them as column identifiers and raises "invalid identifier" errors.4142```sql43-- WRONG: missing colon prefix44SELECT name INTO result FROM users WHERE id = p_id;4546-- CORRECT: colon prefix on both variable and parameter47SELECT name INTO :result FROM users WHERE id = :p_id;48```4950This applies to DECLARE variables, LET variables, and procedure parameters when used inside SELECT, INSERT, UPDATE, DELETE, or MERGE.5152### Semi-Structured Data5354- VARIANT, OBJECT, ARRAY for JSON/Avro/Parquet/ORC.55- Access nested fields: `src:customer.name::STRING`. Always cast with `::TYPE`.56- VARIANT null vs SQL NULL: JSON `null` is stored as the string `"null"`. Use `STRIP_NULL_VALUE = TRUE` on load.57- Flatten arrays: `SELECT f.value:name::STRING FROM my_table, LATERAL FLATTEN(input => src:items) f;`5859### MERGE for Upserts6061```sql62MERGE INTO target t USING source s ON t.id = s.id63WHEN MATCHED THEN UPDATE SET t.name = s.name, t.updated_at = CURRENT_TIMESTAMP()64WHEN NOT MATCHED THEN INSERT (id, name, updated_at) VALUES (s.id, s.name, CURRENT_TIMESTAMP());65```6667> See `references/snowflake_sql_and_pipelines.md` for deeper SQL patterns and anti-patterns.6869---7071## Data Pipelines7273### Choosing Your Approach7475| Approach | When to Use |76|----------|-------------|77| Dynamic Tables | Declarative transformations. **Default choice.** Define the query, Snowflake handles refresh. |78| Streams + Tasks | Imperative CDC. Use for procedural logic, stored procedure calls, complex branching. |79| Snowpipe | Continuous file loading from cloud storage (S3, GCS, Azure). |8081### Dynamic Tables8283```sql84CREATE OR REPLACE DYNAMIC TABLE cleaned_events85 TARGET_LAG = '5 minutes'86 WAREHOUSE = transform_wh87 AS88 SELECT event_id, event_type, user_id, event_timestamp89 FROM raw_events90 WHERE event_type IS NOT NULL;91```9293Key rules:94- Set `TARGET_LAG` progressively: tighter at the top of the DAG, looser downstream.95- Incremental DTs cannot depend on Full-refresh DTs.96- `SELECT *` breaks on upstream schema changes -- use explicit column lists.97- Views cannot sit between two Dynamic Tables in the DAG.9899### Streams and Tasks100101```sql102CREATE OR REPLACE STREAM raw_stream ON TABLE raw_events;103104CREATE OR REPLACE TASK process_events105 WAREHOUSE = transform_wh106 SCHEDULE = 'USING CRON 0 */1 * * * America/Los_Angeles'107 WHEN SYSTEM$STREAM_HAS_DATA('raw_stream')108 AS INSERT INTO cleaned_events SELECT ... FROM raw_stream;109110-- Tasks start SUSPENDED. You MUST resume them.111ALTER TASK process_events RESUME;112```113114> See `references/snowflake_sql_and_pipelines.md` for DT debugging queries and Snowpipe patterns.115116---117118## Cortex AI119120### Function Reference121122| Function | Purpose |123|----------|---------|124| `AI_COMPLETE` | LLM completion (text, images, documents) |125| `AI_CLASSIFY` | Classify text into categories (up to 500 labels) |126| `AI_FILTER` | Boolean filter on text or images |127| `AI_EXTRACT` | Structured extraction from text/images/documents |128| `AI_SENTIMENT` | Sentiment score (-1 to 1) |129| `AI_PARSE_DOCUMENT` | OCR or layout extraction from documents |130| `AI_REDACT` | PII removal from text |131132**Deprecated names (do NOT use):** `COMPLETE`, `CLASSIFY_TEXT`, `EXTRACT_ANSWER`, `PARSE_DOCUMENT`, `SUMMARIZE`, `TRANSLATE`, `SENTIMENT`, `EMBED_TEXT_768`.133134### TO_FILE -- Common Pitfall135136Stage path and filename are **separate** arguments:137138```sql139-- WRONG: single combined argument140TO_FILE('@stage/file.pdf')141142-- CORRECT: two arguments143TO_FILE('@db.schema.mystage', 'invoice.pdf')144```145146### Cortex Agents147148Agent specs use a JSON structure with top-level keys: `models`, `instructions`, `tools`, `tool_resources`.149150- Use `$spec$` delimiter (not `$$`).151- `models` must be an object, not an array.152- `tool_resources` is a separate top-level key, not nested inside `tools`.153- Tool descriptions are the single biggest factor in agent quality.154155> See `references/cortex_ai_and_agents.md` for full agent spec examples and Cortex Search patterns.156157---158159## Snowpark Python160161```python162from snowflake.snowpark import Session163import os164165session = Session.builder.configs({166 "account": os.environ["SNOWFLAKE_ACCOUNT"],167 "user": os.environ["SNOWFLAKE_USER"],168 "password": os.environ["SNOWFLAKE_PASSWORD"],169 "role": "my_role", "warehouse": "my_wh",170 "database": "my_db", "schema": "my_schema"171}).create()172```173174- Never hardcode credentials. Use environment variables or key pair auth.175- DataFrames are lazy -- executed on `collect()` / `show()`.176- Do NOT call `collect()` on large DataFrames. Process server-side with DataFrame operations.177- Use **vectorized UDFs** (10-100x faster) for batch and ML workloads.178179## dbt on Snowflake180181```sql182-- Dynamic table materialization (streaming/near-real-time marts):183{{ config(materialized='dynamic_table', snowflake_warehouse='transforming', target_lag='1 hour') }}184185-- Incremental materialization (large fact tables):186{{ config(materialized='incremental', unique_key='event_id') }}187188-- Snowflake-specific configs (combine with any materialization):189{{ config(transient=true, copy_grants=true, query_tag='team_daily') }}190```191192- Do NOT use `{{ this }}` without `{% if is_incremental() %}` guard.193- Use `dynamic_table` materialization for streaming or near-real-time marts.194195## Performance196197- **Cluster keys**: Only for multi-TB tables. Apply on WHERE / JOIN / GROUP BY columns.198- **Search Optimization**: `ALTER TABLE t ADD SEARCH OPTIMIZATION ON EQUALITY(col);`199- **Warehouse sizing**: Start X-Small, scale up. Set `AUTO_SUSPEND = 60`, `AUTO_RESUME = TRUE`.200- **Separate warehouses** per workload (load, transform, query).201202## Security203204- Follow least-privilege RBAC. Use database roles for object-level grants.205- Audit ACCOUNTADMIN regularly: `SHOW GRANTS OF ROLE ACCOUNTADMIN;`206- Use network policies for IP allowlisting.207- Use masking policies for PII columns and row access policies for multi-tenant isolation.208209---210211## Proactive Triggers212213Surface these issues without being asked when you notice them in context:214215- **Missing colon prefix** in SQL stored procedures -- flag immediately, this causes "invalid identifier" at runtime.216- **`SELECT *` in Dynamic Tables** -- flag as a schema-change time bomb.217- **Deprecated Cortex function names** (`CLASSIFY_TEXT`, `SUMMARIZE`, etc.) -- suggest the current `AI_*` equivalents.218- **Task not resumed** after creation -- remind that tasks start SUSPENDED.219- **Hardcoded credentials** in Snowpark code -- flag as a security risk.220221---222223## Common Errors224225| Error | Cause | Fix |226|-------|-------|-----|227| "Object does not exist" | Wrong database/schema context or missing grants | Fully qualify names (`db.schema.table`), check grants |228| "Invalid identifier" in procedure | Missing colon prefix on variable | Use `:variable_name` inside SQL statements |229| "Numeric value not recognized" | VARIANT field not cast | Cast explicitly: `src:field::NUMBER(10,2)` |230| Task not running | Forgot to resume after creation | `ALTER TASK task_name RESUME;` |231| DT refresh failing | Schema change upstream or tracking disabled | Use explicit columns, verify change tracking |232| TO_FILE error | Combined path as single argument | Split into two args: `TO_FILE('@stage', 'file.pdf')` |233234---235236## Practical Workflows237238### Workflow 1: Build a Reporting Pipeline (30 min)2392401. **Stage raw data**: Create external stage pointing to S3/GCS/Azure, set up Snowpipe for auto-ingest2412. **Clean with Dynamic Table**: Create DT with `TARGET_LAG = '5 minutes'` that filters nulls, casts types, deduplicates2423. **Aggregate with downstream DT**: Second DT that joins cleaned data with dimension tables, computes metrics2434. **Expose via Secure View**: Create `SECURE VIEW` for the BI tool / API layer2445. **Grant access**: Use `snowflake_query_helper.py grant` to generate RBAC statements245246### Workflow 2: Add AI Classification to Existing Data2472481. **Identify the column**: Find the text column to classify (e.g., support tickets, reviews)2492. **Test with AI_CLASSIFY**: `SELECT AI_CLASSIFY(text_col, ['bug', 'feature', 'question']) FROM table LIMIT 10;`2503. **Create enrichment DT**: Dynamic Table that runs `AI_CLASSIFY` on new rows automatically2514. **Monitor costs**: Cortex AI is billed per token — sample before running on full tables252253### Workflow 3: Debug a Failing Pipeline2542551. **Check task history**: `SELECT * FROM TABLE(INFORMATION_SCHEMA.TASK_HISTORY()) WHERE STATE = 'FAILED' ORDER BY SCHEDULED_TIME DESC;`2562. **Check DT refresh**: `SELECT * FROM TABLE(INFORMATION_SCHEMA.DYNAMIC_TABLE_REFRESH_HISTORY('my_dt')) ORDER BY REFRESH_END_TIME DESC;`2573. **Check stream staleness**: `SHOW STREAMS; -- check stale_after column`2584. **Consult troubleshooting reference**: See `references/troubleshooting.md` for error-specific fixes259260---261262## Anti-Patterns263264| Anti-Pattern | Why It Fails | Better Approach |265|---|---|---|266| `SELECT *` in Dynamic Tables | Schema changes upstream break the DT silently | Use explicit column lists |267| Missing colon prefix in procedures | "Invalid identifier" runtime error | Always use `:variable_name` in SQL blocks |268| Single warehouse for all workloads | Contention between load, transform, and query | Separate warehouses per workload type |269| Hardcoded credentials in Snowpark | Security risk, breaks in CI/CD | Use `os.environ[]` or key pair auth |270| `collect()` on large DataFrames | Pulls entire result set to client memory | Process server-side with DataFrame operations |271| Nested subqueries instead of CTEs | Unreadable, hard to debug, Snowflake optimizes CTEs better | Use `WITH` clauses |272| Using deprecated Cortex functions | `CLASSIFY_TEXT`, `SUMMARIZE` etc. will be removed | Use `AI_CLASSIFY`, `AI_COMPLETE` etc. |273| Tasks without `WHEN SYSTEM$STREAM_HAS_DATA` | Task runs on schedule even with no new data, wasting credits | Add the WHEN clause for stream-driven tasks |274| Double-quoted identifiers | Forces case-sensitive names across all queries | Use `snake_case` unquoted identifiers |275276---277278## Cross-References279280| Skill | Relationship |281|-------|-------------|282| `engineering/sql-database-assistant` | General SQL patterns — use for non-Snowflake databases |283| `engineering/database-designer` | Schema design — use for data modeling before Snowflake implementation |284| `engineering-team/senior-data-engineer` | Broader data engineering — pipelines, Spark, Airflow, data quality |285| `engineering-team/senior-data-scientist` | Analytics and ML — use alongside Snowpark for feature engineering |286| `engineering-team/senior-devops` | CI/CD for Snowflake deployments (Terraform, GitHub Actions) |287288---289290## Reference Documentation291292| Document | Contents |293|----------|----------|294| `references/snowflake_sql_and_pipelines.md` | SQL patterns, MERGE templates, Dynamic Table debugging, Snowpipe, anti-patterns |295| `references/cortex_ai_and_agents.md` | Cortex AI functions, agent spec structure, Cortex Search, Snowpark |296| `references/troubleshooting.md` | Error reference, debugging queries, common fixes |