Exasol ai setup

Exasol AI Setup Skill

Trigger when the user mentions notebook-connector, secure config store, Secrets, configure notebook-connector, AI setup, set up credentials for TE, set up credentials for TXAIE, first-time notebook-connector setup, or similar setup tasks.

Purpose

This skill establishes the Python-managed secure configuration that later notebook-connector skills depend on.

After configuration is complete:

• activate exasol-itde for local Docker DB lifecycle
• activate exasol-notebook-connections for Python DB and BucketFS helper APIs
• activate exasol-transformers for Transformers Extension setup and usage
• activate exasol-text-ai for Text AI Extension setup and usage
• activate exasol-bucketfs or exasol-udfs only for deeper BucketFS or SLC work beyond notebook-connector setup

Routing Algorithm

Choose the narrowest path that matches the user request:

1. Python-based configuration

   • Trigger phrases: Secrets, AILabConfig, StorageBackend, save config in python, notebook cell, script
   • Load: references/secrets-python.md
   • Use scripts from: scripts/

2. Validation / smoke tests

   • Trigger phrases: check config, verify connection, validate notebook-connector, smoke test
   • Load: references/validation.md
   • Use scripts from: scripts/

3. Downstream notebook-connector work

   • Trigger phrases: bring_itde_up, open_pyexasol_connection, open_sqlalchemy_connection, open_ibis_connection, open_bucketfs_bucket, initialize_te_extension, initialize_text_ai_extension, deploy_license
   • Hand off to exasol-itde, exasol-notebook-connections, exasol-transformers, or exasol-text-ai after setup validation succeeds

Multiple routes can apply. Load all matching references before responding.

Default Guidance

• Prefer the Python path when the user wants notebook cells, automation, or agent-generated code.
• Before handing off to DB helpers, BucketFS helpers, TE, or TXAIE work, run scripts/validate_config.py to confirm the Secrets store is populated for the expected backend.
• Use AILabConfig as the source of the common key names stored in Secrets.

Validation

After writing or updating config, verify it before handing off to downstream skills:

• run scripts/validate_config.py

Success signals:

• the backend resolves to the expected value (onprem or saas)
• the required setup values for this skill are present in the store so downstream notebook-connector workflows can proceed

Expected failure mode:

• if required setup values for this skill such as storage_backend, db_schema, backend-specific connection values, or real credentials are missing, or the store still contains template values such as my-db-host or placeholder SaaS IDs/PATs, setup validation should fail until the user replaces them with real values

Safety Rules

• Do not guess SaaS account IDs, PATs, passwords, or BucketFS credentials.
• Do not read secrets from config files unless the user explicitly provides them.