Polygres CLI
The public CLI is shipped with PyPI package polygres-cli. Use it for project
setup, migrations, imports, single-row writes, retrieval configuration, API
keys, connection details, and pgContext AI Search collection workflows. Use the
Python SDK for retrieval queries and backend-owned pgContext
automation in application code. If you previously installed the combined
polygres 0.2.x package, follow the package split
migration
before following the workflows below.
CLI setup requires an existing active Polygres account. Signup, organization onboarding, email verification, password management, organization switching, and project deletion remain dashboard workflows.
First standard project from a terminal
Use this workflow when Polygres should host the primary PostgreSQL database. It creates a standard project, applies a migration, configures text retrieval, and creates a Runtime API key. Project creation waits for readiness, while project selection remains an explicit step.
pipx install "polygres-cli==0.4.1"
polygres login
polygres projects create standard "Support Search"
polygres projects use "Support Search"
polygres migrations apply --file ./seed.sql --name seed_documents
polygres text configs create-tsvector documents_body_tsv --table documents --text-column body --generated-column body_tsv --yes
polygres text configs list
polygres keys create local-dev
polygres envWait until text configs list reports ready, save the one-time API-key secret, and use the POLYGRES_RUNTIME_URL line from polygres env with the Python SDK.
First synced project from a terminal
Use this workflow when an existing PostgreSQL database remains the source of truth:
export SOURCE_DATABASE_URL="postgresql://..."
polygres login
polygres projects create sync "Support Search" \
--connection-env SOURCE_DATABASE_URL \
--all-eligible \
--yes
polygres projects use "Support Search"
polygres projects statusFor an interactive setup, run:
polygres projects create sync "Support Search"The CLI securely prompts for the connection URL and table selection. After the project reaches Streaming, configure retrieval over the synchronized tables. Application writes continue through the source PostgreSQL database.
