Techtools Claude Code Cron Loop — Git skill for Claude Code
An autonomous multi-agent system powered by Claude Code running on a cron schedule.
How to install Techtools Claude Code Cron Loop
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open TaraJura/techtools-claude-code-cron-loop and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Techtools Claude Code Cron Loop does
An autonomous multi-agent system powered by Claude Code running on a cron schedule. Agents collaborate through a shared task board, automatically committing all changes to GitHub. - managed by OpenClaw
Alternatives in Git
- My Claude Code Setup — GitHub Repo stars Shared starter template configuration and CLAUDE.md memory bank system Setup template 2.1k ★
- Create Beads Orchestration — Set up lightweight multi-agent orchestration with git-native task tracking for Claude Code 163 ★
- Deep Research — Execute autonomous multi-step research using Gemini Deep Research Agent for market analysis, competitive lands 162 ★
README
Cronloop
**A generic, installable, goal-driven autonomous multi-agent framework.** Set a goal; a pipeline of specialized AI agents (powered by Claude Code) works it on a schedule — turning the goal into tasks, implementing them, verifying them, and improving itself over time. A web **Control Panel** lets you set the goal, edit agent prompts, configure the schedule, and run/monitor everything.
**The Control Panel is now a Rails 8 app (`rails-app/`).** It is a native rewrite of the original Python `control-panel/server.py` + bash `scripts/*.sh`: a Rails 8.1 / Ruby 4 app using Hotwire (Turbo/Stimulus) for the UI, Solid Queue for the agent pipeline and supervisor schedule (replacing the system crontab), and Active Record only for live run state. It still reads/writes the same canonical framework files (`GOAL.md`, `config.json`, `tasks.md`, `actors/*/prompt.md`), so nothing about *how the loop is configured* changes. The Python panel and its docs below remain as the parity reference.
**Run it / promote it to production:** `bash rails-app/deploy/cutover.sh` — idempotent; it bundles, prepares the DBs, precompiles assets, installs the `cronloop-rails` systemd unit (Puma on `127.0.0.1:8788`, Solid Queue in-process), repoints the `ai-brain.cz` nginx vhost from the old `:8787` to `:8788`, and disables the legacy Python service + cron block. Needs (passwordless) sudo for the systemd/nginx steps; without it the script prints those as manual steps. See `rails-app/REWRITE_SPEC.md` for the full design contract.
Nothing here is domain-specific. **To change what the loop builds, edit `GOAL.md`** — the machinery stays the same. The default goal is *self-improvement of this framework*.
- Engine: Claude Code (Anthropic's AI CLI), run headless.
- Dependencies: none. Backend is Python 3 stdlib only; the UI is vanilla JS/CSS (no
build step, no CDNs); scripts are bash +
python3for JSON. Works full
Related Skills
Sweep For Complete
Safe, schedulable sweep that finds started-state issues whose linked PR merged and completes exactly those via
Harnessmesh
Persistent multi-agent collaboration fabric for agentic coding harnesses. Connect Antigravity, OpenAI Codex, C
Automatasaurus Development Guide
Automatasaurus is a workflow orchestration framework for Claude Code that enables multi-agent autonomous devel
Garra Routine
Roadmap-driven autonomous slice. Reads ROADMAP+Linear, picks the next step, plans it, implements via PR + CI,
Factory Agent
Autonomous software factory: Linear issue → Claude Managed Agents → GitHub PR. ~$1 per feature, ~3 minutes.
Outpost
Autonomous GitHub agent powered by OpenCode
Related Agents
Schedules
Configure manual or cron-based Agent runs and inspect each Schedule trigger through SandBase Sessions.
.github/workflows/redteam.yml
name: Continuous Red Team Assessment on: push: branches: [main, develop] pull_request: branches: [main] schedu
Safe Build
Autonomous reversibility-first implementer — executes code/UI changes through the safe-build checkpoint loop (