{"author":"mrtnx","children":[{"author":"chloeeekim","children":[{"author":"mrtnx","children":[],"created_at":"2026-05-01T00:32:48.000Z","created_at_i":1777595568,"id":47970105,"options":[],"parent_id":47960424,"points":null,"story_id":47958608,"text":"&gt; does it actually improve outcomes over time\nYes, after every execution (optionally), you are able to attach self-eval and update the skills accordingly.<p>&gt; does it tend to reinforce suboptimal patterns?\nI was thinking on a ML approach but kept myself lazy and decided go all in with LLM self-evaluation notes.<p>&gt; How much autonomy do the agents actually have in practice?\n100% is possible but if you are asking personally: I define what I need and give them personality&#x2F;mission to accomplish in multiple worksplace.<p>&gt; I\u2019ve found that fully autonomous loops tend to need a lot of guardrails to stay useful.\nI can tell, it did what I was not able to do mentally and physically. It is a huge unlock for creating time for myself.","title":null,"type":"comment","url":null}],"created_at":"2026-04-30T10:18:08.000Z","created_at_i":1777544288,"id":47960424,"options":[],"parent_id":47958608,"points":null,"story_id":47958608,"text":"Interesting approach.<p>I\u2019m especially curious about the \u201cself-learning loop\u201d \u2014 in practice, does it actually improve outcomes over time, or does it tend to reinforce suboptimal patterns?<p>And How much autonomy do the agents actually have in practice?<p>I\u2019ve found that fully autonomous loops tend to need a lot of guardrails to stay useful.","title":null,"type":"comment","url":null},{"author":"fule","children":[{"author":"mrtnx","children":[],"created_at":"2026-05-01T00:34:05.000Z","created_at_i":1777595645,"id":47970112,"options":[],"parent_id":47963959,"points":null,"story_id":47958608,"text":"I assigned it to self-document: <a href=\"https:&#x2F;&#x2F;agentrq.com&#x2F;docs\" rel=\"nofollow\">https:&#x2F;&#x2F;agentrq.com&#x2F;docs</a> I verified all steps one by one. Please let me know if I can help you onboard.","title":null,"type":"comment","url":null}],"created_at":"2026-04-30T15:26:39.000Z","created_at_i":1777562799,"id":47963959,"options":[],"parent_id":47958608,"points":null,"story_id":47958608,"text":"How does a team setup look like? Maybe tested it with someone?","title":null,"type":"comment","url":null}],"created_at":"2026-04-30T05:43:18.000Z","created_at_i":1777527798,"id":47958608,"options":[],"parent_id":null,"points":6,"story_id":47958608,"text":"AgentRQ is a (optionally) human-in-the-loop, self learning closed loop task manager for agents. Agents can create and schedule tasks for themself and work on them on their own schedule.<p>In high level it comes with one supervisor MCP that controls workspaces(worker agents) and unlimited number of isolated workspace MCPs (self learning agents).<p>Each workspace&#x2F;agent has a mission&#x2F;persona for the agent. And self-learning-loop note.<p>I am using it about 6 weeks in production, and completed more than 500 tasks. I just released the opensource version(as is in production) under Apache 2.0 license.<p>Currently it supports Gemini CLI and Claude code. I am going to extend support all major agents soon.<p>Happy to answer any questions.","title":"Show HN: Task Manager for AI Agents (MCP, Opensource)","type":"story","url":"https://github.com/agentrq/agentrq"}
