ironflow sql

Execute a read-only SQL query against the Ironflow database.

ironflow sql <query> [flags]

Queries are automatically scoped to the current environment. Only SELECT statements are allowed. Mutation keywords (INSERT, UPDATE, DELETE, DROP, etc.) are rejected.

Arguments:

Argument	Required	Description
query	Yes	SQL SELECT query to execute

Flags:

Flag	Short	Type	Default	Description
--format		string	table	Output format: table, json, csv, jsonl
--timeout		int	5000	Query timeout in milliseconds
--max-rows		int	1000	Maximum rows to return
--server	-s	string		Server URL override

Available tables: events, runs, steps, functions, projection_registry, projection_state

Environment scoping: Queries are automatically filtered to the current environment via temporary views. The environment is determined by the IRONFLOW_ENV env var, or defaults to default.

Examples:

# Count events by name
ironflow sql "SELECT name, COUNT(*) AS total FROM events GROUP BY name"

# Find failed runs
ironflow sql "SELECT id, status FROM runs WHERE status = 'failed' LIMIT 10"

# Export events as JSON
ironflow sql "SELECT * FROM events WHERE entity_id = 'order-123'" --format json

# CSV output for spreadsheets
ironflow sql "SELECT id, name FROM functions" --format csv

# JSONL for streaming processing
ironflow sql "SELECT * FROM events" --format jsonl | jq '.name'

Output (table):

id              name            entity_id
evt_001         order.placed    order-123
evt_002         order.paid      order-123

---