Skip to main content

AI-powered data generation plugin for sqlseed

Project description

sqlseed-ai

English | 中文

AI-powered data generation plugin for sqlseed.

LLM-driven schema analysis, self-correcting config generation, and template pool assistance. Uses OpenAI-compatible API (default: OpenRouter free models).

Installation

pip install sqlseed-ai

Quick Start

# Set API key (OpenRouter, OpenAI, DeepSeek, etc.)
export SQLSEED_AI_API_KEY="your-api-key"

# Generate AI-suggested YAML config
sqlseed ai-suggest app.db --table users --output users.yaml

# With self-correction (3 rounds by default)
sqlseed ai-suggest app.db --table users --output users.yaml --verify

# Specify model
sqlseed ai-suggest app.db --table users -o users.yaml --model deepseek/deepseek-chat

# Skip cache
sqlseed ai-suggest app.db --table users -o users.yaml --no-cache

Features

Schema Analyzer

SchemaAnalyzer extracts rich context from your database (columns, indexes, sample data, foreign keys, data distribution) and builds a structured prompt for LLM analysis. Returns column-level generation configs as JSON.

Self-Correcting Refiner

AiConfigRefiner validates LLM output against actual schema:

  1. LLM generates column config
  2. Refiner checks for unknown generators, type mismatches, expression errors
  3. If errors found, sends correction request back to LLM
  4. Up to 3 retry rounds, then raises AISuggestionFailedError

Auto Model Selection

Queries OpenRouter API to find the best available free model. Falls back through a priority list:

nvidia/nemotron-3-super-120b-a12b:free → tencent/hy3-preview:free → ...

Result cached for 1 hour. Skip auto-selection by specifying --model or SQLSEED_AI_MODEL.

Template Pool

When sqlseed fills a table with skip_ai=False, the plugin pre-generates candidate values for columns that can't be mapped to a deterministic generator (via sqlseed_pre_generate_templates hook).

File Caching

AI configs cached in .sqlseed_cache/ai_configs/ with schema hash validation. Schema changes auto-invalidate cache. Use --no-cache to skip.

Configuration

Environment Variables

Variable Fallback Default Description
SQLSEED_AI_API_KEY OPENAI_API_KEY API key (required)
SQLSEED_AI_BASE_URL OPENAI_BASE_URL https://openrouter.ai/api/v1 API endpoint
SQLSEED_AI_MODEL auto-select Model name
SQLSEED_AI_TIMEOUT 60 API timeout (seconds)

CLI Options

--model, -m       Model name (overrides auto-selection)
--api-key         API key (overrides env)
--base-url        API base URL (overrides env)
--max-retries     Self-correction rounds (default: 3, 0=disable)
--verify/--no-verify  Toggle self-correction (default: verify)
--no-cache        Skip file cache
--timeout         API timeout in seconds (default: 120)

Plugin Hooks

This plugin registers via [project.entry-points."sqlseed"] and implements:

Hook Purpose
sqlseed_ai_analyze_table LLM-driven table analysis, returns column configs
sqlseed_pre_generate_templates Pre-generate candidate values for complex columns
sqlseed_register_providers Placeholder (no-op, entry-point registration)
sqlseed_register_column_mappers Placeholder (no-op, entry-point registration)

Requirements

  • Python >= 3.10
  • sqlseed >= 0.1.0
  • openai >= 1.0
  • An OpenAI-compatible API key

License

AGPL-3.0-or-later

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sqlseed_ai-0.1.15.tar.gz (18.6 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

sqlseed_ai-0.1.15-py3-none-any.whl (20.8 kB view details)

Uploaded Python 3

File details

Details for the file sqlseed_ai-0.1.15.tar.gz.

File metadata

  • Download URL: sqlseed_ai-0.1.15.tar.gz
  • Upload date:
  • Size: 18.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.13

File hashes

Hashes for sqlseed_ai-0.1.15.tar.gz
Algorithm Hash digest
SHA256 710b14900b051c15091e5155cde6ccdb999405c2891516bd3416f7ec6a04c1bb
MD5 2babb229970f746ffadae4f9df68a3d8
BLAKE2b-256 069a3a592be86b7d3c4c59bf23dbd7b82845f4dfb56d2fbad148638ba9bc3d3f

See more details on using hashes here.

Provenance

The following attestation bundles were made for sqlseed_ai-0.1.15.tar.gz:

Publisher: publish.yml on sunbos/sqlseed

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file sqlseed_ai-0.1.15-py3-none-any.whl.

File metadata

  • Download URL: sqlseed_ai-0.1.15-py3-none-any.whl
  • Upload date:
  • Size: 20.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.13

File hashes

Hashes for sqlseed_ai-0.1.15-py3-none-any.whl
Algorithm Hash digest
SHA256 88b16d9cceedd0112fa47dce8d3a6d50828bc017ae64347c83405e6c2c21590e
MD5 0263938a7f71c85ce595d92b3bde2f4b
BLAKE2b-256 8ea098176ff3c1e7b1aa7621e76eb1870c337e5e018b5b81cff544abc5b951cc

See more details on using hashes here.

Provenance

The following attestation bundles were made for sqlseed_ai-0.1.15-py3-none-any.whl:

Publisher: publish.yml on sunbos/sqlseed

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page