issue-loop

An automated loop for GitHub issues that moves work through specification, implementation, and pull-request review. Each issue and review uses a separate coding-agent session, with people approving design decisions and risky merges.

In plain words
What is it for?
Use it to repeatedly process GitHub issues, limit work by labels or a parent issue, inspect status, run one step at a time, or optionally continue unattended.
Why use it?
It removes the need to start and coordinate a new session for every step while keeping human approval for important decisions.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/nlook-service/issue-template/issue-loop
Any agent
npx skills add nlook-service/issue-template --skill issue-loop
Clone the repo
git clone --depth 1 https://github.com/nlook-service/issue-template

Made for: Claude Code, Codex.

Per session 50 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,180 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00050 $0.07180
Opus 5 $0.00025 $0.03590
Sonnet 5 $0.00010 $0.01436
Haiku 4.5 $0.00005 $0.00718

Measured 2d ago against content hash e2f803a338e2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

issue-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 2d 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.

claude/skills/issue-loop/SKILL.md · 247 lines

How it starts

The opening of the file, as written. The whole thing — 247 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/issue-loop

/spec → /implement-issue → /review-pr을 매번 세션을 열어 돌리는 대신, 루프 드라이버가 GitHub 상태를 읽고 각 걸음을 새 서브에이전트에 위임해 자동 반복한다. 원본의 원칙 유지: ① 이슈 1개 = 세션 1개, ② 리뷰는 구현과 다른 세션 (둘 다 걸음별 독립 서브에이전트로 자동 충족), ③ 설계 분해안·UI 시안 승인과 risk:high 머지는 언제나 사람 — 일반 머지는 기본 모드에선 사람이 결정하고 루프가 실행, 무인 모드에선 강화 리뷰가 대신한다.

Usage

/issue-loop                      # 이어하기: GitHub 상태를 읽고 루프 시작/재개
/issue-loop "<기능 설명>"         # 설계부터: spec(분해안·UI 시안 승인은 사람) → 이슈 등록 → 루프
/issue-loop <상위 이슈 번호>       # 특정 기능([Feature] 이슈)으로 범위 제한
/issue-loop --label <태그>        # 내가 붙인 태그가 있는 이슈만 (ai-task와 AND) — 범위 내 전부 완료가 목표
/issue-loop --once               # 한 걸음만 진행하고 종료
/issue-loop --status             # 상태판 + 다음 계획만 출력 (dispatch 없음)
/issue-loop --max N              # 걸음 수 안전상한 (기본: max(20, 범위 내 이슈 수 × 4))
/issue-loop --unattended         # 완전 무인: agent:auto 이슈만 · 질문 대신 건너뜀 · 강화 리뷰 통과 PR 자동 머지
/issue-loop --economy            # 한도 절약: 구현을 전부 sonnet으로 (리뷰·설계는 opus 유지)
/issue-loop --usage-guard N      # 사용량 추정치가 N% 넘으면 걸음 경계에서 우아하게 중단 (기본 90, ccusage 필요)

0. 사전 확인 (매 실행 첫 단계)

  1. git rev-parse --show-toplevel이 현재 디렉토리와 같은지 — 아니면 대상 저장소 루트로 이동을 안내하고 중단
  2. gh auth status 통과 여부
  3. .claude/commands/implement-issue.md·.claude/commands/review-pr.md 존재 여부 — 없으면 issue-template install.sh 실행을 안내하고 중단 (서브에이전트가 이 파일들을 계약서로 읽는다)
  4. --label <태그>가 주어졌으면: gh label list에 그 태그가 존재하는지, 그 태그가 붙은 열린 ai-task 이슈가 1개 이상인지 확인 — 없으면 오타 가능성을 알리고 중단 (조용히 빈 루프를 돌지 않는다)
  5. 판정 라벨 존재 보장: review:approved·review:rejected·needs-respec는 루프의 상태 원본이다 — gh label list에 없으면 즉시 생성한다 (gh label create <이름> --color <색> 2>/dev/null || true). 없는 채로 돌면 판정 라벨이 안 붙어 같은 PR을 무한 재리뷰한다
  6. 브랜치 보호 확인 (1인 개발 감지): gh api "repos/{owner}/{repo}/branches/<기본브랜치>/protection/required_pull_request_reviews" --jq .required_approving_review_count 2>/dev/null — 값이 1 이상이면, 자기 PR을 자기가 승인할 수 없으므로 1인 리포에선 라벨 승인만으로 gh pr merge가 거부된다고 루프 시작 전에 경고한다 (보호 규칙 완화 / 머지는 사람이 GitHub에서 직접 — 중 택일 안내). 조회 실패(보호 없음 404·권한 부족)는 조용히 건너뛴다
  7. 저널 디렉토리 준비: mkdir -p .claude/issue-loop 하고, .gitignore.claude/issue-loop/가 없으면 추가 (저널은 로컬 상태 — 커밋 금지). 반대로 .design/docs/design/은 커밋 대상이다 — 이슈가 참조하는 계약물이므로 gitignore에 넣지 않는다
  8. 저널 마지막에 ⏸ INTERRUPTED 블록이 있으면 그 내용을 사용자에게 요약해 보여주고 이어서 진행 (상태 원본은 GitHub이므로 저널은 참고 컨텍스트일 뿐 — 재개는 항상 GitHub 상태 재수집으로 시작한다)

Read the full file on GitHub · 247 lines

Changes

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.

  1. 2d ago First seen · 247 lines · 50 tokens per session scan A e2f803a338e2

Subscribe to this mod's changes

issue-loop is a skill published in the GitHub repository nlook-service/issue-template (2 stars, last pushed 1mo ago), licensed MIT. It adds 50 tokens to every session and 7,180 once invoked, about $0.0003 per session on Opus 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 2026-08-31.

Related

Other skills, from other repositories

archify

Create polished, validated architecture, workflow, sequence, data-flow, and lifecycle/state diagrams as explorable standalone HTML with inline SVG, dark/light themes, optional trace motion, and PNG/JPEG/WebP/SVG/WebM export. Accept plain-language requirements or pasted Mermaid flowchart, sequenceDiagram, and…

tt-a1i/archify · 130 tokens

vox-director

Turn ONE topic into a finished Vox-style paper-collage explainer / ad video, end to end on the Atlas Cloud API + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Use this whenever the user wants a "Vox style" video, a paper/torn-paper collage animation, a "motion collage"…

Alisa0808/vox-director · 236 tokens

social-post

學習使用者的 Facebook/Instagram/YouTube/Threads/X 語氣與受眾,規劃、撰寫、確認後發佈內容;以已登入 Chrome 受控掃描、草擬及回覆 FB/IG/Threads 留言;並作為流量、留存與轉化的結構化帳本。使用者說「發文」「用我的口氣」「回覆留言」「自動回留言」「掃留言」「查流量」「演算法」「把數據訓練進去」「比較貼文」「優化 pattern」時使用。.

Hao0321/claude-skill-social-post · 135 tokens

privacy-jp

日本サービス向け(일본 서비스용) プライバシーポリシー・利用規約・同意モーダル・Cookieバナー 자동 생성. 個人情報保護法(APPI)·消費者契約法·特定商取引法 반영. Next.js 1316 프로젝트 대상.

kimlawtech/korean-privacy-terms · 74 tokens

higgsfield

Use this skill whenever the user asks anything about Higgsfield AI — writing or refining video/image prompts, choosing a model (Kling, Sora 2, Veo, Wan, Seedance, Minimax Hailuo, DoP, Soul, Nano Banana, Seedream, Flux, GPT Image, etc.), camera controls, named motion presets, Soul ID character consistency, Cinema…

OSideMedia/higgsfield-ai-prompt-skill · 147 tokens

higgsfield-camera

Use when the user asks about camera movements, shot types, or how to describe camera behavior in a Higgsfield prompt. Contains all named camera controls with descriptions, best use cases, and example prompt phrases.

OSideMedia/higgsfield-ai-prompt-skill · 47 tokens