Skip to content
Home › CLI › Flags

Common flags

A handful of flags appear on multiple commands. Listing them once here so you don't have to memorise them per command. Each flag below works on /run, /apply, /run-apply, /sync, and /profile unless otherwise noted.

Prerequisites

  • AMX installed.
  • An active DB and LLM profile (run /setup if not).

Flag reference

Flag Type Default What it does
--db-profile <name> string active profile Override the active DB profile for this command only
--llm-profile <name> string active profile Override the active LLM profile for this command only
--scope <kind> database/schema/table/column inferred Force the scope picker level instead of inferring from positional args
--profiling-mode <mode> full/sampled/metadata profile default Override the profiling mode for one run
--profiling-sample-size <n> int profile default (5000) Override sample size when --profiling-mode sampled
--review-all flag off Force the review wizard to walk every row, not just low-confidence ones
--auto-accept-high flag on Bulk-accept high-confidence drafts. Pair with --review-all to override
--apply flag off Auto-/apply after /run finishes. Skips the review wizard for high+medium
--csv <path> string — Write the run's review queue to a CSV file in addition to (or instead of) the in-memory store
--csv-only flag off Skip the in-memory store; writes only to the --csv path
--debug flag off Verbose logging — every SQL statement, every LLM call, every retry
--skip-network flag off Skip the connectivity check at command start (useful when /doctor already ran)
--dry-run flag off On /apply: print the SQL that would run; don't execute
--limit <n> int — On /run: limit to the first N tables in scope. Useful for incremental runs
--filter <expr> string — On /run: restrict to tables whose name matches the regex
--llm-batch-size <n> int profile default (10) Override column_batch_size for one run

Common combinations

Sweep a whole warehouse cheaply

amx /run \
  --profiling-mode metadata \
  --llm-profile openai-mini \
  --auto-accept-high \
  --csv ~/amx-output/sweep-2026-05-03.csv

metadata mode skips row scans entirely (no warehouse cost), --llm-profile openai-mini uses the cheap model, --auto-accept-high lets the review wizard skip ahead through high-confidence rows, and the CSV gives you a portable record outside the AMX history store.

Re-run only the tables that failed last time

amx /run \
  --filter '^(stg_|fct_|dim_)' \
  --limit 50 \
  --debug

The regex matches your dbt-style staging / fact / dimension tables; --limit 50 keeps the run bounded; --debug surfaces the exact failure if it happens again.

Apply without the review wizard

amx /run-apply \
  --auto-accept-high \
  --review-all=false \
  --apply

/run-apply is /run immediately followed by /apply. --auto-accept-high accepts high-confidence drafts; everything else is queued for later interactive review.

Do not pair --apply with a freshly-tuned LLM profile

--apply skips the review wizard for high+medium confidence drafts. Until you've confirmed the LLM profile produces good descriptions for your schema, leave the wizard interactive. Re-add --apply only after one or two successful manual sweeps.

Test profiling cost without LLM calls

amx /profile sales.customer \
  --profiling-mode full \
  --skip-network

/profile runs only the Profile agent — it samples the table and prints stats but doesn't call the LLM. Use it to estimate how long full-mode profiling would take on your warehouse before unleashing /run.

Sample config — defaults that act like flags

Most flags have a per-profile default that lives in ~/.amx/config.yml:

db_profiles:
  prod-pg:
    backend: postgresql
    # ...
    profiling_mode: sampled            # → --profiling-mode default
    profiling_sample_size: 5000        # → --profiling-sample-size default

llm_profiles:
  openai-prod:
    provider: openai
    model: gpt-4o
    column_batch_size: 10              # → --llm-batch-size default
    n_alternatives: 3
    temperature: 0.2
    logprob_high: 0.85
    logprob_medium: 0.50

Set sensible defaults per profile so you only need flags for one-off overrides.

What's next