The Owner Can Be an Agent

Codex "cache forever" isn't working for me. Same thread resumed after 13h: 0% cached. ~94K-token thread, codex exec, ChatGPT login, 6.1 Sol: 2h: 99.8% 13h: 0% 25h: 8% (system prompt probably got cached) Desktop too: https://github.com/openai/codex/issues/56050

the models get lost, the app keeps breaking and stalling my goals, they write giant unreadable .txt reports, which i open in this tiny markdown panel which crushes these 5 column tables into 150px. like come on

Hey I really recommend not using goal. You’re much better off with a manager agent paired with a scheduled automation that reminds it to poll progress of different runner agents it starts in separate threads. My pro tip after that, is to tell the manager to write all your steers down, and reconcile them against the overall goal before dispatching new work.

hey! this is a gap in our repo lifecycle that we'll work on fixing ASAP. in the meantime if you DM me your account info or support ticket # i can fix this for you now!

Tasks and Elicitations are available today in the Claude Code CLI, rolling out to more Claude products soon. Building plugins for Claude? Add Tasks and Elicitations to your MCP server today. https://code.claude.com/docs/en/mcp#mcp-tasks-for-server-authors https://code.claude.com/docs/en/mcp#respond-to-mcp-elicitation-requests

A word to the wise if you are on a personal @Grok @bot account and thinking of turning it into a team account: be aware of one thing. I recently onboarded a colleague to Grok bot and converted to a team account by inviting them from my personal account. Cursor makes it seem easy, and once they accept, you have a team account. I thought it was great—until I realized my personal Origin account with my code repositories did not transfer automatically. That broke my development flow. The bots under the team account can no longer access the codebase still under my personal account. My account ID changed, and there is no user-driven way to migrate the codebase. Support confirmed they must handle it manually. With high ticket volume, it will take three to four days. I switched on Wednesday but cannot do production work until they transfer it. This is not urgent for me, but it is extremely annoying and poor design. For professional developers it could be disruptive. There was no warning of this risk. If you convert a personal account to a team account and keep your code in Origin, you will likely lose access until support manually transfers it. You have been warned.

Copilot CLI 1.0.102: subagents now honor the custom and MCP tool approvals you already gave, so no duplicate permission prompts. 1.0.103 reuses cached local MCP tool snapshots across worktrees in one repo. Good time to review what you've approved before subagents inherit it.

Agents SDK v0.30.0 now supports Pi 1.4 with smarter hibernation and crash recovery. Your harnesses can recover from evictions and storage errors automatically. https://developers.cloudflare.com/changelog/post/2026-10-22-pi-harness-pi-1-4/

The last few days with my Hermes setup have been ROUGH. Broken Kanban handbacks left me manually coordinating agents that were supposed to work autonomous. I finally repaired handoffs and an isolated coding pilot. That’s progress, but the real test is whether useful work finishes without me babysitting it. Ugh!

fwiw in the most recent versions of paperclip we've curbed this too-many-subtasks behavior

How different is this from @papercliping ? I’ve used it in the past, eventually the agents go crazy with their own sub-tasks and it becomes more complex than it should. How are you guardrailing these?

Ya'll could just use one agent, Hermes, and be done with this madness lol

Day 16/ You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.

Claude Code 2.1.304 adds decider: "person" to mod tool checks. Only the permission dialog or a host callback can approve that call. With nobody around to ask, it's refused with the mod's reason. I'd put deploys and prod DB writes behind it and let auto mode handle the rest.

the whole agent is composed of tasks. each task is made of phases. a phase is: store intent durably, execute effect, store result + next phase. process dies. reopen, get all previously live tasks and rerun the last phase they've been in. no event sourcing. that doesn't work well with long running sessions.

Copilot CLI 1.0.100 (Oct 19): its sandbox now lets commands reach localhost and private-network services by default. That's where local MCP servers, dev databases and admin panels live. If you enabled the sandbox to contain the agent, re-check its reach. https://github.com/github/copilot-cli/releases/tag/v1.0.100

I think most people multi-tasking with agents could benefit from slowing down most of the stuff you're doing wont really matter- what's your P0? what would it look like if you gave it your full attention?

Andrzej, how much control does Eve give you over retries around tool calls? If an 'add to cart' call succeeds just before a worker crashes, resuming the run shouldn't add the item twice.

I've been building ecommerce agents for a while and tried many AI frameworks along the way. I must say: @eve is the first one that really clicked. I told @shardara this recently and he asked me why, so I thought it was worth sharing publicly. In my first AI consulting gigs (mostly custom shopping assistants) I was stunned by how much infra work it takes to ship an agent. I'd say 60–70% of my time went into infra, and only the rest into the actual AI work: context engineering, data, evals. That was insane tbh, because none of that infra felt custom to me. Most frameworks either didn’t help much with infra, or if they did, they forced clunky abstractions on the AI side, and I kept fighting the framework instead of doing my work. That's probably why so many folks say "don't use a framework, just build from scratch." I felt that too (we all have our Claudes and Codexes, right?), but going framework-less never felt right either. There’s so much repeatable stuff, and way too many footguns, to rebuild it every time. So for a long time I was looking for something very simple: 1. Pls pls handle infra for me. Durable execution, sandboxes, streaming, reconnects, subagents, file uploads that land in the sandbox, auth, web search, easy deployment. 2. LEAVE ME FREEDOM on the AI part. I want to own the harness, not have the framework own it for me. @eve is just the first one that ticks all these boxes. Bonus: every production agent quickly needs a UI and often a custom API. A playground, a context view, a domain-specific data viewer (inspired by @HamelHusain, highly recommend). With Eve + Next.js it’s all in a single repo (not even a monorepo!), deployed with one click. Framework-defined Infrastructure at its finest, it really feels good. btw, I come from front-end world and I remember exactly how the world looked before and after nextjs. Now it feels like this again, just with AI. Everything we do at the eCommerce AI Lab at @commerce_ui now runs on Eve: shopping assistants and analytics agents for our clients, plus our own company agent that lives in Slack. It all just works, and a huge amount of custom code just evaporated. Highly recommend

🦞 OpenClaw v2026.10.5 is out! 🤖 Sidebar keeps your agents within reach ↩️ Skill Workshop 2.0 gets Undo 🔗 Public conversation links need no sign-in 🎉 2,250 PRs and 637 contributors. Just the highlights. Much more in the full notes 👇 https://docs.openclaw.ai/releases/2026.10.5

I've had a few people ask why I'd ever bother creating Kvarm when solutions like Cursor Origin exist. It's simple. We need an open-source git platform for the agentic era. No vendor lock-in.

harrison would say that :) you also want to take into account kv cache and dreaming optimizations that only harness-model colocation can achieve. ive had my own career ruined by trying to be switzerland without realizing the costs. truth is his advice probably: - low end: doesnt matter - mid tier: maybe - high end: definitely and most people are not high end (>2T tokens a month)

I WAS WRONG! I am so sorry. Skill issue. My fault. It had a memory saved from earlier this year when I was setting up auth and I told it not to login for me since I needed to test it myself. It saved it as a hard rule and I only found it because I had Codex comb through CLAUDE.md and all of Claude Code’s memory files. Once Codex tweaked those memories, it was able to do it. It seems a lot of people experienced something similar so I suggest that you audit its memory and see if there are any restrictions it might’ve set on itself to be unintentionally strict. I am once again humbled by the consequences of my own actions.

kind of sad I have a pi durable coding agent with durable subagents, but I'm not using it because I wanna use opus55, and its long subagent-using sessions completely crumble on a restart

Code Review is now generally available for Claude Team and Enterprise plans. Reviews are now faster, catch more of the bugs that matter, and admins can pick from Quick, Standard, or Deep review levels for each repo.

Origin repositories build, preview, and deploy on Netlify the same way GitHub, GitLab, and Bitbucket repos do. Push a commit, and Netlify starts a build. Open a pull request, and the deploy preview link shows up right in the PR. It's in beta through Cursor's invite-only Origin program.

Can Cursor Origin give each cloud agent its own permissions? I’d like an agent to push fixes and open PRs without being able to merge just because my account can. And I’d want every action linked to the exact agent run.