opengeni

Explain OpenGeni session-based agent service architecture and client API usage.

47|5|Updated Apr 16, 2026
One-click install
npx skills add https://github.com/Cloudgeni-ai/opengeni --skill opengeni
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: opengeni
Source: https://github.com/Cloudgeni-ai/opengeni/tree/main/.agents/skills/opengeni
Command: npx skills add https://github.com/Cloudgeni-ai/opengeni --skill opengeni

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps you orient yourself and make accurate changes when working with OpenGeni’s evolving, session-based agent service—so you can avoid stale documentation and find the real contracts in code.

Core Features & Use Cases

  • Architecture and boundaries: Clarifies how the API, durable event log, Temporal orchestration, worker activities, sandbox backends, object storage, and MCP tools fit together.
  • Client integration guidance: Helps design external clients that create sessions, stream/replay events, upload files, attach resources, handle approvals/interrupts, and manage scheduled tasks.
  • Source-anchored debugging workflow: Guides you to discover endpoints, public shapes, env/config, persistence semantics, and runtime behavior via targeted repository searches.
  • Safety-first claims discipline: Promotes careful wording and verification so you don’t incorrectly assert auth, reliability, retries, or tool/mount behavior that isn’t current.

Quick Start

Ask the AI to inspect the current OpenGeni repo and explain how a client should create a session, stream SSE events, and replay missing events after reconnect.

Frequently Asked Questions about opengeni

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create an OpenGeni session and stream SSE events from a client?

To integrate an external client with OpenGeni sessions, inspect the repository code to discover current endpoint shapes for creating sessions, streaming SSE events, and replaying missing events after reconnects. This source-anchored approach avoids outdated documentation.

What is the architecture boundary between Temporal orchestration and sandbox backends?

The architecture boundary separates Temporal orchestration, which manages durable worker activities, from sandbox backends, which execute MCP tools and file resource flows. Inspecting the codebase reveals how these components interact with the API and object storage without relying on stale docs.

How do I upload files and attach resources to an agent session?

To upload files and attach resources to an agent session, query the repository's API and object storage code to verify current persistence semantics and resource flow contracts. This ensures your client integration handles attachments safely according to the actual implementation.

Does OpenGeni support MCP tool integration within sandbox backends?

OpenGeni supports MCP tool integration within sandbox backends, but you must inspect the worker runtime and sandbox tooling code to confirm current tool and mount behavior. This verification-safe approach prevents incorrectly asserting tool execution semantics.

Why does my OpenGeni deployment have incorrect retry or auth behavior?

Incorrect retry or auth behavior often stems from relying on outdated documentation instead of current code. You should debug by inspecting the repository's env/config settings, Temporal orchestration, and event log semantics to verify actual runtime contracts.

What's the best way to debug OpenGeni's scheduled tasks and approval interrupts?

The best way to debug scheduled tasks and approval interrupts is using a source-anchored workflow to inspect the Temporal orchestration and API boundary code. This reveals the actual persistence semantics and event flow contracts for handling interrupts safely.