Atmos is an infrastructure runtime that coordinates tools such as Terraform, OpenTofu, Kubernetes, Helm, Packer, Ansible, and containers through consistent commands and configuration. It is for teams running cloud infrastructure on laptops, in CI, or through AI agents across environments and regions. Its catalogue entries provide skills, agents, commands, and other add-ons for Atmos workflows.
Borrowing it
Nothing to install: this file belongs to cloudposse/atmos. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/cloudposse/atmos/main/.claude/skills/pr-maintenance-loop/SKILL.mdgit clone --depth 1 https://github.com/cloudposse/atmosWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/cloudposse/atmos/pr-maintenance-loop)<a href="https://agentmods.dev/skills/cloudposse/atmos/pr-maintenance-loop"><img src="https://agentmods.dev/badge/skills/cloudposse/atmos/pr-maintenance-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cloudposse/atmos/pr-maintenance-loop"><img src="https://agentmods.dev/badge/skills/cloudposse/atmos/pr-maintenance-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00123 | $0.01416 |
| Opus 5.5 | $0.00049 | $0.00566 |
| Sonnet 5.5 | $0.00025 | $0.00283 |
| Haiku 4.5 | $0.00012 | $0.00142 |
Grade A, and why
pr-maintenance-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hourly PR Maintenance Loop
Schedules the fix-all skill to run every hour via Claude Code's native
/loop primitive (CronCreate/ScheduleWakeup) — not GitHub Actions, no new external service,
no secrets beyond the local git/gh session already in use. All the actual check/fix logic
(sync, CI, CodeRabbit threads, lint, coverage, code-hygiene — the security model and audible
notifications that govern it) lives in fix-all; this skill only owns the scheduling.
Known limitations (surface these, don't hide them)
- Session-only, and there's no cloud-compatible alternative. The loop lives in this Claude
Code process, using this exact worktree and this local
gh/gitsession (signing key, gh auth, checkout state) — that's the whole point, per the intro above ("no new external service, no secrets beyond the local git/gh session already in use"). It dies when the process ends and auto-expires after 7 days. It only fires while idle. There is no cross-session persistence — that's why this skill re-establishes the loop every session rather than assuming one is already running. A cloud-scheduled agent runs in a fresh, isolated environment with none of that local state, so it cannot run this particular loop — there's no cloud path to fall back to. That's why Step 1 below callsCronCreatedirectly instead of going through the generic/loopskill:/loopwould ask whether to use a cloud schedule for anything ≥60 minutes, and that question has no useful answer here. - Codex parity is unsolved. This skill only covers Claude Code sessions. If the user runs a Codex workspace on the same branch, tell them explicitly that this loop does not cover it.
Step 0 — preconditions
gh pr view --json number,state,mergeStateStatusfor the current branch. If there's no open PR, tell the user and stop — don't create one.- Check whether a loop is already running this session (list scheduled jobs; look for the
sentinel first line
[PR-MAINTENANCE-LOOP pr=#<number> repo=cloudposse/atmos started=<ts>]). If found for this PR, don't start a second one — report it's already active. A fresh session never inherits a stale job, since the old one died with the old process, so this only guards against double-starting within one session.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 96 lines · 123 tokens per session scan A c89db81fad30
pr-maintenance-loop is a skill published in the GitHub repository cloudposse/atmos (1,403 stars, last pushed today), licensed Apache-2.0. It adds 123 tokens to every session and 1,416 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
git
Git and code hosting platform management — GitHub, GitLab, Codeberg, Forgejo. Zero barrier — install x-cmd and manage repos from the terminal. Use for "git", "github", "gitlab", "repo", "pull request", "code hosting".
vhs-e2e-gif
Record a test run, a TUI session, or any terminal command as a GIF with VHS and attach it to a GitHub PR as a release-hosted asset, never a repo commit. Use when asked to record an e2e run, demo a fix on a PR, attach a GIF or screen recording to a pull request, show a test passing visually, or produce a terminal…
apply-qa-fixes
Apply QA gate findings then hand back for re-review. Never self-approves or closes. Use when: apply QA fixes, remediate gate FAIL/CONCERNS, /apply-qa-fixes.
aiox-qa
Test Architect & Quality Advisor (Quinn). Use for comprehensive test architecture review, quality gate decisions, and code improvement. Provides thorough analysis including requ...
hatch3r-refactor
Internal code quality improvement workflow without changing external behavior. Use when refactoring code structure, simplifying modules, or improving maintainability.
looper-qa
Use when a Looper-managed GitHub repo needs scheduled pre-merge QA — a PR carries the looper:qa label, the spec stage reaches looper:spec-ready, or the Looper reviewer loop requests an independent second pass. Runs the full QA cycle (Looper state probe → PR checkout → ffs code review → language-specific test suite →…