jentic-apitools-pipelines 0.0.0a7

Jentic Apitools Pipelines - Orchestrate score, repair, and storage into workflow pipelines

pipelines

High-level workflows that orchestrate analyze, score, repair, and storage operations for OpenAPI specifications.

Key Features

The pipelines package provides two main workflow functions: import_openapi runs the full import pipeline (fetch, parse, validate, convert, bundle, repair, score, store) and score_openapi runs a lighter scoring pipeline (fetch, parse, bundle, validate, score). Both functions accept an OASJsonRequest with processing configuration and an optional progress callback for real-time updates. The package handles format conversion from Swagger 2.0 and Google Discovery to OpenAPI 3.x, reference bundling via Redocly, and artifact management through configurable storage contexts. The repo_tools module provides rebuild_scores_json and rebuild_apis_json for rebuilding the root scores.json and apis.json catalog files from all API specs in a local repository clone.

Dependencies

Internal: common, llm, analyze, score, repair, storage. External: jentic-openapi-transformer (with redocly), datadog.

Installation

uv add pipelines

Quick Start

from jentic.apitools.common.models import OASJsonRequest, OASRequestMeta, OASProcessConfiguration
from jentic.apitools.pipelines import score_openapi

request = OASJsonRequest(meta=OASRequestMeta(oas_process_configuration=OASProcessConfiguration()))
result = score_openapi(request, spec_url="https://petstore3.swagger.io/api/v3/openapi.json")
print(f"Success: {result.success}")

Testing

uv run --package pipelines pytest packages/pipelines/tests -v