A working blueprint that shows how the GitHub platform delivers full CI/CD for an Oracle APEX application — from idea to production, with gates, traceability, and a recovery story.

A 15-minute customer demo that walks an APEX change from issue → branch → pull request → merge → release → production, with a recovery story when something breaks. The technical pipeline uses SQLcl + GitHub Actions to deploy schema changes and APEX application updates to Oracle Autonomous Database.

GitHub platform surfaces shown:

.github/
  workflows/
    validate.yml         # PR static checks + deploy plan comment
    deploy.yml           # main → DEV; release → PROD (gated)
  ISSUE_TEMPLATE/        # feature_request, bug_report
  pull_request_template.md
  CODEOWNERS
apex/
  f100.sql               # APEX app export (placeholder; replace with real export)
  README.md
db/
  install.sql            # master orchestrator
  schema/
    000_schema_migrations.sql
    001_create_feedback.sql
  migrations/
    002_seed_data.sql
    003_add_category.sql # THE DEMO MOMENT
  README.md
scripts/
  deploy.sql             # SQLcl entry point
  verify.sql             # post-deploy smoke
docs/
  RESEARCH.md            # full technical research blueprint
  SETUP.md               # one-time setup (ADB, wallet, secrets, environments)
  BUILD_APEX_APP.md      # A→Z walkthrough for replacing synthetic f100.sql with a real app
  DEMO_RUNBOOK.md        # operational runbook for the "add category field" demo
  DEMO_SCRIPT.md         # 15-minute act-by-act speaker notes

If you're here to run the demo:

1. Read docs/SETUP.md — provision Oracle Autonomous DB Always Free, encode your wallet, configure GitHub Secrets per environment.
2. (Optional) docs/BUILD_APEX_APP.md — build and export a real APEX app to replace the synthetic placeholder.
3. docs/DEMO_RUNBOOK.md — step-by-step procedure for staging and executing the "add category field" demo end-to-end (copy-pasteable commands, browser tab order, on-stage narration cues, reset procedure).
4. docs/DEMO_SCRIPT.md — higher-level 15-minute act-by-act narrative for adapting the demo to other changes.

If you're here to understand the pipeline:

• Start at docs/RESEARCH.md — the full technical blueprint with citations.
• Then read .github/workflows/deploy.yml — the actual deploy pipeline (~150 lines).

1. Stakeholder opens issue: "Add 'category' field to feedback form".
2. Developer branches, adds db/migrations/003_add_category.sql and the matching APEX page change, opens a PR.
3. validate.yml runs on the PR: lints workflows, checks SQL hygiene (WHENEVER SQLERROR EXIT), enforces migration ordering, blocks any committed secrets, and posts a deploy-plan PR comment.
4. Merge → DEV : deploy.yml runs deploy-dev, applies migrations, imports the APEX app, verifies.
5. Tag a Release (v1.1.0) → deploy-prod waits for a human reviewer on the production environment → approved → applied to PROD.
6. Rollback : a bad commit gets caught by the pipeline; git revert PR restores green.

Schema migrations are plain SQLcl. To smoke-test locally without GitHub:

# 1. Install SQLcl (https://www.oracle.com/database/sqldeveloper/technologies/sqlcl/download/)
# 2. Place your wallet zip at ./wallet/ (unzipped), chmod 600 ./wallet/*
# 3. export TNS_ADMIN=$PWD/wallet
# 4. sql USER/PASS@yourdb_high
SQL> @scripts/deploy.sql
SQL> @scripts/verify.sql
SQL> apex import -applicationid 100 -workspacename DEMO -file apex/f100.sql

The following are intentionally out of scope to keep the demo at 15 minutes. See docs/RESEARCH.md for production-grade upgrade paths.

• Liquibase migrations
• Codespaces / devcontainer
• Multi-workspace / per-PR ADB clones
• utPLSQL / Cypress tests
• Static security scanning (APEX-SERT)
• Self-hosted runners

This is a demo repository. No production support. Adapt freely.