Skip to content

cli-usage.md

Latest commit

 

History

History
111 lines (70 loc) · 2.81 KB

cli-usage.md

File metadata and controls

111 lines (70 loc) · 2.81 KB

CLI Usage

Commands live in backend/app/cli.py and share services with the HTTP API where noted.

How to run

Mode Command
Local (venv) ./scripts/cli-local.sh --help
Docker ./scripts/cli-docker.sh --help

Run from repository root so paths and env files resolve correctly.

Data pipeline

ingest <path>

  • Partial-accept ingestion of a PropFlux JSON array.
  • Persists ingestion_jobs, raw_listings, listings, and rejected_listings.
  • Prints job_id, status, and valid/invalid counts.

score <job-id>

  • Runs advanced_v2 batch scoring into score_results for the job’s listings.

validate-dataset <job-id>

  • Dataset-level quality gates; persists validation row and writes backend/output/job_<id>_validation.json.

analyze <job-id>

  • Analytics stage for the ingestion job (marks job analyzed).

export <job-id> --format json|csv

  • Exports scored job results to output/.

inspect-rejections <job-id>

  • Summarizes rejection reasons for a job (--top-fields, --sample-values).

Ranking (Week 3+ / API parity)

rank-query

Executes the same ranking path as POST /api/v1/rankings/query.

./scripts/cli-local.sh rank-query \
  --dataset-source job:1 \
  --strategy-preset rental_income \
  --top-n 10

Common options: --province, --city, --suburb, --price-min, --price-max, --property-type, --bedrooms-min, --bathrooms-min, --confidence-min, --page, --page-size, --weight-override signal=0.25, --output-json path.json.

listing-detail

./scripts/cli-local.sh listing-detail --run-id <run_id> --listing-id <id> --pretty

profiles-list / profile-show

./scripts/cli-local.sh profiles-list
./scripts/cli-local.sh profile-show --preset rental_income --pretty

Evaluation and performance

evaluate-scoring

Runs scoring evaluation gates (promote/revert/experimental) per config/scoring.yaml thresholds.

benchmark-baseline

Writes performance baseline artifacts under backend/output/ (includes API latency measurements where configured).

Example: full local sequence

./scripts/migrate.sh
./scripts/cli-local.sh ingest backend/tests/fixtures/propflux/mixed_valid_invalid.json
./scripts/cli-local.sh score 1
./scripts/cli-local.sh validate-dataset 1
./scripts/cli-local.sh rank-query --dataset-source job:1 --strategy-preset rental_income --top-n 5

Or use the bundled demo script: demo.md → ./scripts/demo-local.sh.

Ingestion job statuses

Status Meaning
created Job created, not started
processing Pipeline running
completed All records valid
completed_with_errors Partial accept
analyzed Analytics stage done
failed Fatal failure