Releases·Tool Release
Developer open-sources Toren, a self-hosted runtime for long-running background AI agents
A developer released Toren, an open-source runtime for AI agents that run for hours or days entirely on self-hosted machines. Agents are defined as folders containing a YAML config, markdown instructions, and TypeScript tools, with Postgres as the only dependency. The Apache 2.0 release targets practitioners who want long-lived agents without cloud services or telemetry.
4 engagement1 sourceFri, Aug 28, 2026, 03:44 PM
AI Overview
- Agents are defined as a folder: a YAML file with model and limits, a markdown instructions file, and TypeScript tools the agent can call
- Postgres is the only dependency; there is no cloud version, no account, and no telemetry — nothing phones home
- Agents can get a sandboxed Docker container for bash access, their own Telegram bot, and cron schedules
- Human approval gates can be configured for actions the operator does not want the agent doing alone
- Licensed Apache 2.0; built over several months and shared on Reddit on 2026-08-28 with low early engagement
Entities
Related
Products·launched
Meowbert: self-hosted open-source agent platform running each task in its own Docker container
Meowbert is a self-hosted, open-source agent platform where LLM-driven tasks run autonomously in isolated Docker containers. It targets people doing long-running projects who want a clean local alternative to hosted agent apps, working with any model.
Meowbert is a self-hosted, open-source agent platform where LLM-driven tasks run autonomously in isolated Docker containers. It targets people doing long-running projects who want a clean local alternative to hosted agent apps, working with any model.
9 engagement1 sourceredditThu, Oct 1, 2026, 03:51 PM
Products·launched
Rebuno - open-source execution runtime for operating AI agents in production
Rebuno is an open-source runtime that manages execution state and guardrails for AI agents in production, regardless of the framework they were built with. Agents submit tool and model calls as steps before executing them, and Rebuno evaluates each step against per-agent YAML policies, allowing, denying, or holding calls for human approval. It solves the problem of safely operating agents built with different frameworks in production environments.
Rebuno is an open-source runtime that manages execution state and guardrails for AI agents in production, regardless of the framework they were built with. Agents submit tool and model calls as steps before executing them, and Rebuno evaluates each step against per-agent YAML policies, allowing, denying, or holding calls for human approval. It solves the problem of safely operating agents built with different frameworks in production environments.
3 engagement1 sourcehackernewsMon, Sep 14, 2026, 04:29 PM
Releases·Tool Release
Cloudgeni open-sources Opengeni (Apache-2.0), an agent infrastructure layer for sessions, sandboxes, approvals, and cost tracking
Cloudgeni released Opengeni, an open-source layer of agent infrastructure extracted from two years of internal rebuilding: crash-resilient sessions, tool-call sandboxes, human approval gates, and per-run cost tracking. It is model-agnostic (works with Claude and open models), already runs in enterprises with 30,000 employees, and can be self-hosted free or used hosted at model cost plus 5%. The launch went live on Product Hunt with the founder answering questions.
Cloudgeni released Opengeni, an open-source layer of agent infrastructure extracted from two years of internal rebuilding: crash-resilient sessions, tool-call sandboxes, human approval gates, and per-run cost tracking. It is model-agnostic (works with Claude and open models), already runs in enterprises with 30,000 employees, and can be self-hosted free or used hosted at model cost plus 5%. The launch went live on Product Hunt with the founder answering questions.
1 engagement1 sourceredditMon, Oct 5, 2026, 03:20 PM
Releases·Tool Release
Developer launches GratefulAgents open-source platform for running coding agents in Kubernetes
A developer built GratefulAgents, an open-source platform that runs coding agents in isolated Kubernetes pods to offload resource-intensive tasks from local machines. The Go-based runtime supports tools, MCP, approvals, retries, handoffs, asynchronous subagents, context compaction, memory, usage tracking, and tracing, with support for Claude and OpenAI models.
A developer built GratefulAgents, an open-source platform that runs coding agents in isolated Kubernetes pods to offload resource-intensive tasks from local machines. The Go-based runtime supports tools, MCP, approvals, retries, handoffs, asynchronous subagents, context compaction, memory, usage tracking, and tracing, with support for Claude and OpenAI models.
1 engagement1 sourceredditMon, Jul 20, 2026, 11:43 AM
Products·building
Durable runtime for agent swarms with crash recovery
A runtime that makes agent swarms resilient to crashes by decoupling planning from execution and using an append-only event log for durable execution. It solves the problem of losing all progress when a sandboxed agent crashes, targeting developers building multi-agent systems.
A runtime that makes agent swarms resilient to crashes by decoupling planning from execution and using an append-only event log for durable execution. It solves the problem of losing all progress when a sandboxed agent crashes, targeting developers building multi-agent systems.
5 engagement1 sourceredditTue, Jul 21, 2026, 04:10 AM