Skip to content
Back to skills

Loops

ASecurity

Repeating / autonomous work cycles: bounded loop mode ("loop 10 30m <prompt>"), unattended night-shift or heartbeat work with one-task-at-a- time focus, and a library of proven loop prompts. Use when the user asks for autonomous repeated execution — "keep going until done", "work in cycles", "[HEARTBEAT]", "while I sleep", cron-style self-runs. One contract; the parts under parts/ carry the mechanics.

  • 15 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 7, 2026
ai-agentsgonode

Security analysis

A100/100

Pro scans all 10 files and shows the line behind each finding

Scanned November 7, 2026

npx -y skills add youlianvr/oper-share --skill loops --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Loops?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Loops
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/youlianvr-loops/badge)](https://www.skillsdirectory.com/skills/youlianvr-loops)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: loops
description: >-
  Repeating / autonomous work cycles: bounded loop mode ("loop 10 30m
  <prompt>"), unattended night-shift or heartbeat work with one-task-at-a-
  time focus, and a library of proven loop prompts. Use when the user asks
  for autonomous repeated execution — "keep going until done", "work in
  cycles", "[HEARTBEAT]", "while I sleep", cron-style self-runs. One
  contract; the parts under parts/ carry the mechanics.
---

# loops — autonomous cycles

> Home branch in Memora: `root/skill/loops`.
>
> Every loop must have a stopping condition and a verification step per
> iteration. A loop without a verify step is a way to burn tokens, not work.

## Choosing a part

| Situation | Part |
|---|---|
| Run a bounded repeating cycle from a user prompt | `parts/loop/` |
| Finish an assigned unattended goal with checkpoints and valid stops | `parts/autonomous-work/` |
| Ready-made loop prompt patterns to pick from | `parts/loop-library/` |

## Rules

- **The stop condition belongs to the owner.** Iteration cap and exit condition
  come from the invocation; the agent may propose one, never widen or replace it.
- **Night shift may**: read, test, edit code in its own branch, prepare the
  report. It may not deploy, spend money, or take external actions — those wait
  for the owner, one authorization each.
- **Heartbeat stays silent** when nothing changed: write only on a change or a
  blocker. No progress chatter, no forced "still working" lines.
- Verify changed work and save checkpoints (Memora node + short chat line) so a
  resumed session continues instead of repeating.
- Destructive or irreversible actions inside a loop always pause for the owner;
  autonomy never outranks Never Destroy.
- One-off persona/style modes (`godmode`, `bro`) are separate top-level skills,
  not loops.

## Related

`autonomous-work`, `task-cycle`, `planning`, `ship` (deploy needs the owner).

Files in this skill

  • SKILL.md1.9 KB
  • parts/autonomous-work/PROPOSALS.md6.3 KB
  • parts/autonomous-work/SKILL.md3.2 KB
  • parts/loop-library/LICENSE1 KB
  • parts/loop-library/README.md2 KB
  • parts/loop-library/SKILL.md12.8 KB
  • parts/loop-library/agents/openai.yaml249 B
  • parts/loop-library/references/audit.md2.7 KB
  • parts/loop-library/references/discover.md3.8 KB
  • parts/loop/SKILL.md8.3 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.