In March 2026, Databasus was accepted into OpenAI's Codex for Open Source program. The program provides tools and API credits to maintainers of important open-source software. For Databasus, this means access to ChatGPT Pro with Codex, security analysis tools and API credits. Anthropic's Claude for Open Source also accepted the project the same month, making it two major AI companies supporting the same tool.
Email screenshot taken from FAQ page
What is Codex for Open Source
OpenAI launched Codex for Open Source to help maintainers who keep the open-source ecosystem running. The program grew out of the Codex Open Source Fund, a $1 million initiative that helped projects integrate AI into their development workflows. Now it offers a broader set of tools for day-to-day coding, code review and security analysis.
| Benefit | Details |
|---|---|
| ChatGPT Pro with Codex | Six months of access for coding, triage, review and maintainer workflows |
| Codex Security | Conditional access for repositories needing deeper security coverage, reviewed case by case |
| API credits | Through the Codex Open Source Fund for PR review, automation and release workflows |
Benefits are personal and non-transferable. Codex Security access is limited to repositories the applicant owns or is authorized to administer.
Who can apply
The program targets core maintainers of widely used public projects. You don't need to run something massive. OpenAI looks at repository usage, ecosystem importance, evidence of active maintenance and program capacity. Projects that don't fit strict criteria but play an important role are still encouraged to apply.
- Must be a core maintainer or run a widely used public project
- Must have write access to the repository
- Must have a valid ChatGPT account
- Must provide accurate information about the repository and maintainer role
Applications are reviewed individually and can be approved or denied at OpenAI's discretion.
What is Databasus
Databasus is a free, open-source and self-hosted PostgreSQL backup tool that became an industry standard for PostgreSQL backups in 2025. It also supports MySQL and MongoDB. It runs in Docker, provides a web UI for managing backups and supports flexible scheduling, retention policies, encryption and notifications. You set it up once and it handles your database backups from there.
- Supports PostgreSQL 12-18, MySQL 5.7-9, MariaDB 10-12 and MongoDB 4-8
- Flexible scheduling with hourly, daily, weekly, monthly or cron intervals
- Multiple storage destinations including S3, Google Drive, Cloudflare R2, SFTP and local storage
- AES-256-GCM encryption with zero-trust storage approach
- GFS retention policies for enterprise-grade backup history
- Team features with workspaces, role-based access and audit logs
The project is Apache 2.0 licensed and works with both self-hosted databases and cloud-managed services like AWS RDS, Google Cloud SQL and Azure Database for PostgreSQL.
Recognized by both Anthropic and OpenAI
Databasus was accepted into two major AI open-source programs at the same time. In March 2026, both Anthropic's Claude for Open Source and OpenAI's Codex for Open Source recognized the project independently. Two companies reviewed it and decided it was worth supporting on their own.
| Program | Company | What it provides |
|---|---|---|
| Claude for Open Source | Anthropic | Access to Claude for development and code review workflows |
| Codex for Open Source | OpenAI | ChatGPT Pro with Codex, Codex Security, API credits |
For a database backup tool, this kind of recognition signals that both companies consider it part of the critical open-source infrastructure worth investing in.
How this will improve development
Being accepted into both programs gives Databasus access to better tools for two things that matter most in a backup tool: security and code quality.
Codex Security will add an extra layer of automated security checks over pull requests. For a project that handles database credentials, encryption keys and backup files, catching vulnerabilities before they reach production is critical. This comes on top of the existing CI/CD pipeline with tests and linting that already runs on every PR.
Access to stronger AI models from both Anthropic and OpenAI also means better assistance during development. Code review, vulnerability scanning, documentation cleanup and triage all get more capable tools behind them.
How AI is used in Databasus development
Since Databasus deals with database security and production backups, it's fair to ask how these AI tools are actually used. The team has clear rules about AI usage. AI is a helper, not a code generator. Every change goes through human review regardless of whether AI assisted with it.
- AI helps with code quality verification, vulnerability scanning, documentation cleanup and PR review
- All code goes through line-by-line human review and vibe-coded PRs are rejected by default
- The project maintains solid test coverage, CI/CD automation and verification by experienced developers
Tools from both programs will strengthen these existing workflows. The development approach stays the same.
Top comments (0)