OpenAutonomyX builds autonomous intelligence systems that decide, explain, and act with governance, observability, and human review where it matters.

OpenAutonomyX repositories are organized around:

• Autonomous agents and agent platforms
• Governed decision systems
• Identity, policy, and access-aware automation
• Data, taxonomy, and entity-resolution workflows
• Documentation, review, and reusable instruction assets

Shared execution guidance lives in openautonomyx/common-instructions.

Use that repository for:

• Shared prompt and agent-role guidance
• Engineering execution standards
• Review, audit, and documentation prompts
• Context, guardrail, testing, and air-gapped operation policies
• Reusable instruction boilerplates

Do not duplicate shared instruction policies across product repositories. Product repositories should reference the shared layer and add only repo-specific guidance.

• common-instructions: reusable prompts, standards, validators, policies, and execution guidance
• org-docs: organization-level vision, strategy, architecture, and ADRs
• shared-assets: reusable non-code assets
• shared-repos: reusable code, configuration, tooling, and packages

• Keep repository purpose explicit.
• Preserve current, next, and future state in docs.
• Prefer reviewable, minimal changes over broad undocumented rewrites.
• Require reviewer approval and HITL sign-off for production-facing changes.
• Keep GitHub as the system of record for approved prompts, docs, reviews, and releases.