Thanks for contributing.
ODEI is not a generic AI app collection. It is a research and infrastructure stack built around constitutional world models, local-first memory, governed writes, and auditable execution.
Check that your contribution fits one of these lanes:
- research and architecture
- API and runtime infrastructure
- MCP and agent integrations
- world model tooling and memory systems
- documentation that clarifies the system contract
If a proposal only adds marketing noise, speculative token language, or product sprawl without architectural justification, it is unlikely to land.
- fix a bug with a clear reproduction path
- improve documentation around real system behavior
- tighten type safety or test coverage
- improve operational visibility, guardrails, or auditability
- add examples grounded in actual ODEI APIs or MCP flows
- describe the architectural intent, not only the code diff
- keep changes scoped and legible
- include tests when behavior changes
- call out operational risk when touching public or production surfaces
- avoid unrelated churn
Strong issues include:
- the affected repository or surface
- current behavior
- expected behavior
- reproduction steps or concrete evidence
- logs, screenshots, or payloads when relevant
Do not file public issues for vulnerabilities or secret exposure.
Use:
- GitHub Security Advisories where available
- security@odei.ai
For docs and integration questions, start with api.odei.ai/docs.
For research context, use odei.ai.