fe-git-operator

fe-git-operator is an agent for Claude Code from sh5623/fe-rail. It costs 55 tokens per session (3,122 once invoked), scanned A, original, MIT.

A Git operations agent that organizes changes into separate commits, stages only named files, and writes Conventional Commits messages. Git is the version-control system used to track code changes; a commit is a saved group of related changes.

In plain words
What is it for?
Use it after implementation to split changes into logical commits, choose the package manager for checks, write commit bodies, and safely stage and push the work.
Why use it?
It keeps unrelated work from being mixed together and avoids broad staging commands that may include sensitive or unintended files. Destructive Git commands are forbidden.

Agent for Claude Code

Written for Claude Code: effort in frontmatter. Also seen: model in frontmatter; mentions CLAUDE.md.

Part of the fe-rail plugin — 5 skills, 15 agents, 4 hooks shipped together

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 agents/sh5623/fe-rail/fe-git-operator
Clone the repo
git clone --depth 1 https://github.com/sh5623/fe-rail

Made for: Claude Code.

Or install fe-rail, the plugin that ships this one along with the rest of its 5 skills, 15 agents, 4 hooks.

Wrote 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.

agentmods badge for fe-git-operator

README.md
[![agentmods](https://agentmods.dev/badge/agents/sh5623/fe-rail/fe-git-operator.svg)](https://agentmods.dev/agents/sh5623/fe-rail/fe-git-operator)
Your own site
<a href="https://agentmods.dev/agents/sh5623/fe-rail/fe-git-operator"><img src="https://agentmods.dev/badge/agents/sh5623/fe-rail/fe-git-operator.svg" alt="Measured on agentmods" height="20"></a>
Per session 55 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,122 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 5.1 $0.00055 $0.03122
Opus 5 $0.00028 $0.01561
Sonnet 5 $0.00011 $0.00624
Haiku 4.5 $0.00006 $0.00312

Measured 5d ago against content hash 6d13e2230dd9, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

fe-git-operator 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 5d 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.

agents/fe-git-operator.md · 285 lines

How it starts

The opening of the file, as written. The whole thing — 285 lines — stays where its author put it; the contents beside it link to each section on GitHub.

fe-git-operator Agent

안전한 커밋 분리·스테이징·푸시 전담 에이전트입니다.


목표:

  • 논리 단위별 커밋 분리와 컨벤셔널 커밋 메시지 작성
  • git add ./-A 없이 파일을 명시적으로 스테이징
  • 파괴적 git 명령 실행 없이 안전하게 작업

사용 시점:

  • fe-start Phase 6-1 — 구현 완료 후 커밋 단계 (push·PR 생성은 fe-pr-author 담당)
  • 변경 파일이 여러 논리 단위로 분리 가능한 경우
  • 민감 파일 포함 여부 확인이 필요한 경우

Persona

  • [Identity] git 히스토리를 아름답게 유지하는 꼼꼼한 커밋 큐레이터
  • [Mindset] 커밋 하나 = 논리 단위 하나. 크고 모호한 커밋은 만들지 않는다
  • [Communication] 무엇을 스테이징했고, 어떤 커밋을 만들었는지 표로 보고

패키지 매니저 감지

PM="npm"; PX="npx"
[ -f "pnpm-lock.yaml" ] && PM="pnpm" && PX="pnpm"
[ -f "yarn.lock" ]      && PM="yarn" && PX="yarn"
{ [ -f "bun.lockb" ] || [ -f "bun.lock" ]; } && PM="bun"  && PX="bun"

$PX — 바이너리 직접 실행 (tsc 등) / $PM — npm 스크립트 실행 (lint 등)

커밋 메시지 형식 (Conventional Commits)

<type>: <제목>                 ← scope 없는 경우 (일반, 50자 권장 / 72자 한도, 명령형, 마침표 없음)
<type>(<scope>): <제목>      ← scope 있는 경우 (모노레포·범위 명확화에 한해 — 예: fix(auth):)

<본문>                        ← 한 줄 띄우고 시작, 72자에서 줄바꿈, "무엇을·왜" 설명

<푸터>                        ← (선택) BREAKING CHANGE / Closes #N
규칙 설명
type 화이트리스트 feat / fix / refactor / style / docs / test / chore / perf / ci
제목 길이 50자 권장, 72자 한도
제목 형식 명령형 동사("추가"·"수정"), 마침표 없음, 한국어 가능
scope 선택 — 모노레포·범위 명확화에 한해 사용 (예: fix(auth):)
본문 fix·feat·perf·breaking 은 필수, 그 외 자명한 변경은 생략 가능 — "무엇을·왜"를 쓰고 "어떻게"는 코드에 맡긴다
푸터 호환성 깨짐 → BREAKING CHANGE: <설명> / 이슈 연결 → Closes #N

type 분류 기준

type 사용 시점
feat 새 기능, 새 컴포넌트
fix 버그 수정
refactor 동작 변경 없는 코드 구조 개선
style 포맷팅, CSS 변경
docs 문서 수정 (CLAUDE.md, README 등)
test 테스트 추가·수정
chore 설정, 의존성, 빌드 관련
perf 성능 개선
ci CI/CD 설정

본문 작성 — type별 필수 내용

핵심 원칙: 버그는 "왜 틀렸고 어떻게 고쳤는지", 기능은 "무엇이 생겼고 어떻게 동작하는지"를 본문이 스스로 설명한다. 6개월 뒤 이 커밋만 보고도 맥락이 복원돼야 한다. 추측·일반론 금지 — 실제 diff에 근거해 작성한다.

fix — 버그 수정 (증상 → 원인 → 해결)

Read the full file on GitHub · 285 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. 5d ago First seen · 285 lines · 55 tokens per session scan A 6d13e2230dd9

Subscribe to this mod's changes

fe-git-operator is an agent published in the GitHub repository sh5623/fe-rail (1 stars, last pushed 25d ago), licensed MIT. It adds 55 tokens to every session and 3,122 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 agents, from other repositories

stitch-kit

Use this agent for anything Stitch-related: ideating designs through conversation, generating UI screens from text, editing/iterating designs, generating design variants, managing Stitch Design Systems, converting designs to production code, extracting design tokens, and running the full design-to-ship pipeline.…

gabelul/stitch-kit · 237 tokens

planner

Reads feature spec/design docs, creates PLAN.md and CONTEXT.md inside the feature directory, then spawns the lead-engineer agent after user confirmation. Invoked by the /dev skill. Supports solo mode (default) and team mode (/dev --team) with domain analysis and task tagging.

ByeongminLee/nextjs-claude-code · 61 tokens

security-reviewer

Audits code for security vulnerabilities against SECURITYSTRATEGY.md. Checks OWASP Top 10, secrets, injection, auth gaps, data exposure. Invoked by /security or conditionally by /review.

ByeongminLee/nextjs-claude-code · 46 tokens

spec-writer

Writes or updates feature spec.md and design.md files for Next.js/React projects. Clarifies missing information before writing. Never modifies source code. Invoked by the /spec skill.

ByeongminLee/nextjs-claude-code · 41 tokens

codebase-analyzer

Deep code analysis agent for legacy projects. Reads an external project folder (read-only), extracts feature logic, data flows, business rules, and implicit requirements. Outputs a structured ANALYSIS.md. Invoked by reforge-orchestrator during Phase 1.

ByeongminLee/nextjs-claude-code · 56 tokens

create-orchestrator

Orchestrates the /create ideation pipeline. 5 phases — context gathering, forcing questions (user's language), approach generation, C-level review (CEO/CTO/CPO/CMO), decision & spec conversion. Writes to spec/create/[name]/. Invoked by the /create skill.

ByeongminLee/nextjs-claude-code · 67 tokens