RUMBO IA · AGENT RELIABILITY

Terms of use

These terms apply to the current skills-only RUMBO Agent Reliability release made available for use on OpenAI surfaces. Listing status and availability are determined by OpenAI's own surfaces; these terms do not independently claim endorsement.

Scope

The plugin provides five evidence-first instruction skills for recovering canonical state, reconciling evidence, controlling bounded goal loops, verifying observed effects, and auditing final state. The initial release does not include a RUMBO-operated MCP server, remote tool, account connection, billing flow, write action, or production-control authority.

User responsibility

The plugin is intended to improve reasoning and verification workflows, not to replace required human approvals, access controls, professional judgment, or independent verification. Retrieved context and generated recommendations do not create authority to deploy, spend funds, modify external systems, or perform consequential actions.

Availability and changes

Capabilities, availability, compatibility, and these terms may change as the product evolves. Future releases that add remote services, external accounts, or materially different data handling may be governed by updated terms and privacy disclosures.

Third parties and warranties

OpenAI, ChatGPT, Codex, and other third-party names belong to their respective owners. Any current listing, review, or availability status is determined solely by OpenAI's own surfaces; these terms do not independently represent endorsement or guarantee continued availability. The plugin is provided without a guarantee that it will detect every stale state, contradiction, execution failure, or unsupported claim.

Contact

Questions: sebastian@rumbo.verso.fans.

Last updated: 13 September 2026.

\n \n