Reported News
Agent Autoloop 1.0 runs Claude Code and Codex in the background, asking only when you must decide
Agent Autoloop, an MIT-licensed open-source project, tagged v1.0.0 on GitHub on November 3, 2026. It runs command-line agents such as Claude Code and Codex in the background from schedules and events, and its README says "Any CLI tool that takes a prompt and prints a final reply can plug in."
- Changes reported
- 6
- listed below, each from a source we read
- We tested
- Not tested yet
- reported from sources; we did not run it ourselves
- Gaps we flag
- 3
- things we could not verify
- Sources
- 2
- 2 primary, linked below
What changed
| # | What the source says | Status |
|---|---|---|
| 01 | The repository was created on November 3, 2026 and v1.0.0 ("Agent Autoloop 1.0.0") was tagged the same day; it installs with one curl command on macOS or Linux and needs Node 24, Git and a signed-in CLI agent. | Reported |
| 02 | Rules in ~/.agent-autoloop/rules.json map a cron schedule, a GitHub webhook or an HTTP event to an agent and a working folder; built-in agent plugins cover Claude Code, Codex, ACP and any generic command. | Reported |
| 03 | An agent works one round and exits; when it needs a decision it appends an agent-autoloop JSON block, and the hub sends a card to desktop notifications, Feishu, DingTalk, WeCom, Slack or Discord, then resumes the same conversation after you answer. | Reported |
| 04 | Safety rules from the README: external text such as issue bodies never goes into the prompt, events carrying cwd, agent or access fields are rejected, and dispatch stops after 8 hops. | Reported |
| 05 | The README reports a kernel of 1,651 lines with zero runtime dependencies, 88.78 ms cold start and 2,394 durable writes per second, measured on an Apple M5 Pro. | Reported |
| 06 | A delivery plugin takes a GitHub PR labelled deliver through review, CI, merge, deploy and a live-version check, without checking out PR code on the worker machine. | Reported |
reported by the sourcemeasured or run by usnot verified
Primary source: Agent Autoloop English README on GitHub. We are the fast, independent index; the vendor’s own page is the source of record for this release.
In the source’s words
“Any CLI tool that takes a prompt and prints a final reply can plug in.”
“Past 8 hops, the kernel records the event, stops dispatching, and sends you a notice card.”
“CLI output and default notification strings are currently in Chinese.”
What we have not verified
- We did not install Agent Autoloop or run an agent through it.
- The benchmark numbers are the project's own, measured on one Apple M5 Pro machine; we did not reproduce them.
- The README says the DingTalk, WeCom, Slack and Discord templates were checked only against local mock servers, and the one-line commands for other agents (Cursor, pi, Hermes and others) have not all been run by the project.
This brief is Reported: written from the sources below, not from running the release ourselves.
Sources
- Agent Autoloop English README on GitHubprimary source · fetched 2026-11-04
- Agent Autoloop v1.0.0 releaseprimary source · fetched 2026-11-04
Related
Spotted an error? See our corrections policy or email the address on the contact page. Markdown version for agents: /news/agent-autoloop-1-0-background-cli-agents.md
Want these briefs in your Google Top Stories? Add Agentic Weekly as a preferred source on Google.
Newsletter
The Weekly Verdict. Free, every Monday.
A free email every Monday: the week's few big AI releases, then 5 short use cases for the new models, each with a prompt you can paste.
Free. Double opt-in. One click to unsubscribe.