jenkins-job-insight

Introduce the service, its main inputs and outputs, supported AI providers, and the primary user workflows.

Map the FastAPI app, analyzer, Jenkins client, Jira integration, storage, HTML reporting, examples, and tests so readers can understand how the system is composed.

List prerequisites including Python 3.12, Jenkins access, and external AI CLI authentication, then cover local package installation.

Show how to start the service with the packaged entrypoint or uvicorn and verify it through /health, /docs, and OpenAPI.

Document the provided Dockerfile and docker-compose setup, persistent /data storage, provider credentials, and OpenShift-specific runtime considerations.

Walk through submitting a first analysis request and reading the returned JSON, result URL, and HTML report.

Document every environment variable and matching request override for Jenkins, AI, callbacks, Jira, SSL, logging, and timeouts.

Explain Claude, Gemini, and Cursor CLI authentication and the provider-agnostic execution model used by the service.

Describe Jenkins connection settings, optional repository cloning for source-code context, allowed repository URL schemes, and custom prompt resolution via raw_prompt or JOB_INSIGHT_PROMPT.md.

Explain Cloud and Server/DC authentication, automatic enablement rules, project scoping, search behavior, and AI-based relevance filtering.

Cover POST /analyze, sync versus async execution, callbacks, polling, successful-build handling, and recursive child-job analysis.

Show how to use POST /analyze-failures with either structured failures or raw JUnit XML and explain enriched_xml output.

Document the bundled pytest example that posts junitxml output to the service and rewrites the XML artifact with AI annotations.

Explain lazy HTML report generation, refresh behavior, grouped root-cause cards, result browsing, and dashboard search and pagination.

Explain how identical failures are grouped by signature, how failed child jobs are discovered, and how summaries and counts are computed.

Show how PRODUCT BUG analyses carry search keywords and matched Jira issues to reduce duplicate bug filing.

Summarize endpoints, job states, absolute URLs in responses, and the generated OpenAPI and Swagger interfaces.

Reference the Jenkins-analysis request and response schema, sync flag, background job behavior, and callback fields.

Reference direct failure and raw_xml inputs, validation rules, response shapes, and enriched XML behavior.

Reference GET /results, GET /results/{job_id}, GET /results/{job_id}.html, GET /dashboard, GET /health, and related supporting endpoints.

Document request and response models such as AnalyzeRequest, AnalyzeFailuresRequest, AnalysisResult, FailureAnalysis, ChildJobAnalysis, and JiraMatch.

Describe SQLite persistence, default DB and report paths, pending/running/completed/failed states, and HTML report caching.

Explain forwarded-header handling and how result and report URLs are constructed when the service runs behind proxies or ingress.

Help operators diagnose Jenkins auth failures, missing AI configuration, XML parse errors, Jira issues, callback failures, and report generation problems.

Cover the pytest suite, pre-commit hooks, linting, typing, secret scanning, and the repository layout for maintainers.