Adoption

Agent Skills are supported by leading AI development tools.

VS Code Gemini CLI GitHub Goose Amp Cursor Claude Code Letta OpenCode Claude OpenAI Codex Factory VS Code Gemini CLI GitHub Goose Amp Cursor Claude Code Letta OpenCode Claude OpenAI Codex Factory

junghan0611/commit

Name: commit
Author: junghan0611

skills/commit/SKILL.md

npx skillsauth add junghan0611/agent-config commit

Clean

TrivyContainer and dependency vulnerability scanner

Clean

SemgrepStatic code analysis for vulnerabilities

Clean

mcp-scan (Snyk)Model Context Protocol security validation

Skipped

Snyk (dep)Open source security scanning

Skipped

Socket.devSupply chain security analysis

Skipped

VirusTotalMulti-engine malware detection

Skipped

CrowdStrikeAdvanced threat intelligence

Skipped

OSV-ScannerOpen Source Vulnerability database check

Skipped

OWASP Dep-Check

Create a git commit for the current changes using a concise Conventional Commits-style subject.

Format

<type>(<scope>): <summary>

type REQUIRED. Use feat for new features, fix for bug fixes. Other common types: docs, refactor, chore, test, perf.
scope OPTIONAL. Short noun in parentheses for the affected area (e.g., api, parser, ui).
summary REQUIRED. Short, imperative, <= 72 chars, no trailing period.

Notes

Body is OPTIONAL. If needed, add a blank line after the subject and write short paragraphs.
Do NOT include breaking-change markers or footers.
Do NOT add sign-offs (no Signed-off-by).
Do NOT add AI attribution (no Generated with Claude, Co-Authored-By: Claude, etc.). Keep the log clean.
Only commit; do NOT push.
If it is unclear whether a file should be included, ask the user which files to commit.
Treat any caller-provided arguments as additional commit guidance. Common patterns:
- Freeform instructions should influence scope, summary, and body.
- File paths or globs should limit which files to commit. If files are specified, only stage/commit those unless the user explicitly asks otherwise.
- If arguments combine files and instructions, honor both.

Steps

Infer from the prompt if the user provided specific file paths/globs and/or additional instructions.
Review git status and git diff to understand the current changes (limit to argument-specified files if provided).
(Optional) Run git log -n 50 --pretty=format:%s to see commonly used scopes.
If there are ambiguous extra files, ask the user for clarification before committing.
Stage only the intended files (all changes if no files specified).
Run git commit -m "<subject>" (and -m "<body>" if needed).

Post-commit — agenda stamp (required)

The commit is the agent's job; push is GLG's. Once a commit is pushed, stamp it so the agenda link resolves. Do not stamp local-only commits — the link may break.

# 1. Collect commit info
REMOTE=$(git remote get-url origin)
REPO_URL=$(echo "$REMOTE" | sed -E 's|git@github(-[a-z]+)?\.com:|https://github.com/|;s|\.git$||')
REPO_NAME=$(basename "$REMOTE" .git)
REPO_TAG=$(echo "$REPO_NAME" | sed 's/[-.]//g')   # strip hyphens/dots: homeagent-config → homeagentconfig
SHA=$(git rev-parse --short HEAD)
MSG=$(git log -1 --pretty=%s)

# 2. Agenda stamp (with commit link)
SCRIPT="$HOME/.pi/agent/skills/pi-skills/agenda/scripts/agenda-stamp.sh"
[ -x "$SCRIPT" ] || SCRIPT="$HOME/.claude/skills/agenda/scripts/agenda-stamp.sh"  # per-harness fallback
"$SCRIPT" \
  "${REPO_NAME}: ${MSG} [[${REPO_URL}/commit/${SHA}][${SHA}]]" \
  "pi:commit:${REPO_TAG}"

Optional — Google Chat notification (one CLI call, no token cost):

source ~/.env.local && gog chat messages send "$GOG_CHAT_SPACE_ID" \
  --account "$GOG_CHAT_ACCOUNT" \
  --text "🔨 *${REPO_NAME}* commit: ${MSG}
→ ${REPO_URL}/commit/${SHA}"

Notes:

Multiple sequential commits → stamp only the last one.
Env vars (GOG_CHAT_*) live in ~/.env.local (see PRIVATE.md).
If agenda-stamp.sh fails after reasonable retries, STOP and report the exact command + error. Never fall back to Write/Edit/heredoc on the same target.
For release tags (not commits), the stamp uses pi:release: — see the tag-release skill.

junghan0611/commit

skills/commit/SKILL.md

Read this skill before making git commits

6 stars

tools

Updated Jun 1, 2026

$ install --global

skillsauth

npx skillsauth add junghan0611/agent-config commit

Install this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.

Security Scan Results

3 of 9 scanners reported clean

Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.

Scanners Passed

Scanners in report

Clean

TrivyContainer and dependency vulnerability scanner

95%

Clean

SemgrepStatic code analysis for vulnerabilities

95%

Clean

mcp-scan (Snyk)Model Context Protocol security validation

95%

Skipped

Snyk (dep)Open source security scanning

50%

Skipped

Socket.devSupply chain security analysis

50%

Skipped

VirusTotalMulti-engine malware detection

50%

Skipped

CrowdStrikeAdvanced threat intelligence

50%

Skipped

OSV-ScannerOpen Source Vulnerability database check

50%

Skipped

OWASP Dep-Check

50%

Last scanned: Jun 1, 2026, 2:13 AM220.7s1 file scanned

SKILL.md

name:: commit
description:: Read this skill before making git commits

Create a git commit for the current changes using a concise Conventional Commits-style subject.

Format

<type>(<scope>): <summary>

type REQUIRED. Use feat for new features, fix for bug fixes. Other common types: docs, refactor, chore, test, perf.
scope OPTIONAL. Short noun in parentheses for the affected area (e.g., api, parser, ui).
summary REQUIRED. Short, imperative, <= 72 chars, no trailing period.

Notes

Body is OPTIONAL. If needed, add a blank line after the subject and write short paragraphs.
Do NOT include breaking-change markers or footers.
Do NOT add sign-offs (no Signed-off-by).
Do NOT add AI attribution (no Generated with Claude, Co-Authored-By: Claude, etc.). Keep the log clean.
Only commit; do NOT push.
If it is unclear whether a file should be included, ask the user which files to commit.
Treat any caller-provided arguments as additional commit guidance. Common patterns:
- Freeform instructions should influence scope, summary, and body.
- File paths or globs should limit which files to commit. If files are specified, only stage/commit those unless the user explicitly asks otherwise.
- If arguments combine files and instructions, honor both.

Steps

Infer from the prompt if the user provided specific file paths/globs and/or additional instructions.
Review git status and git diff to understand the current changes (limit to argument-specified files if provided).
(Optional) Run git log -n 50 --pretty=format:%s to see commonly used scopes.
If there are ambiguous extra files, ask the user for clarification before committing.
Stage only the intended files (all changes if no files specified).
Run git commit -m "<subject>" (and -m "<body>" if needed).

Post-commit — agenda stamp (required)

The commit is the agent's job; push is GLG's. Once a commit is pushed, stamp it so the agenda link resolves. Do not stamp local-only commits — the link may break.

# 1. Collect commit info
REMOTE=$(git remote get-url origin)
REPO_URL=$(echo "$REMOTE" | sed -E 's|git@github(-[a-z]+)?\.com:|https://github.com/|;s|\.git$||')
REPO_NAME=$(basename "$REMOTE" .git)
REPO_TAG=$(echo "$REPO_NAME" | sed 's/[-.]//g')   # strip hyphens/dots: homeagent-config → homeagentconfig
SHA=$(git rev-parse --short HEAD)
MSG=$(git log -1 --pretty=%s)

# 2. Agenda stamp (with commit link)
SCRIPT="$HOME/.pi/agent/skills/pi-skills/agenda/scripts/agenda-stamp.sh"
[ -x "$SCRIPT" ] || SCRIPT="$HOME/.claude/skills/agenda/scripts/agenda-stamp.sh"  # per-harness fallback
"$SCRIPT" \
  "${REPO_NAME}: ${MSG} [[${REPO_URL}/commit/${SHA}][${SHA}]]" \
  "pi:commit:${REPO_TAG}"

Optional — Google Chat notification (one CLI call, no token cost):

source ~/.env.local && gog chat messages send "$GOG_CHAT_SPACE_ID" \
  --account "$GOG_CHAT_ACCOUNT" \
  --text "🔨 *${REPO_NAME}* commit: ${MSG}
→ ${REPO_URL}/commit/${SHA}"

Notes:

Multiple sequential commits → stamp only the last one.
Env vars (GOG_CHAT_*) live in ~/.env.local (see PRIVATE.md).
If agenda-stamp.sh fails after reasonable retries, STOP and report the exact command + error. Never fall back to Write/Edit/heredoc on the same target.
For release tags (not commits), the stamp uses pi:release: — see the tag-release skill.

Related Skills

junghan0611/tag-release

documentation

VerifiedTrustedCommunity

Cut an OpenClaw-style CalVer tag for a repo — gather commits, refresh CHANGELOG, promote, tag, push, stamp. Use when the user says '태그 박자', '릴리즈 컷', 'changelog 정리', 'cut a release/tag', or wants a vYYYY.M.D[-suffix] snapshot.

6SKILL.mdUpdated Jun 1, 2026

junghan0611/tag-release

junghan0611/recall

development

VerifiedTrustedCommunity

멀티하네스용 /recall 래퍼. native custom command surface가 없는 하네스(예: Antigravity, Codex)에서도 /recall에 해당하는 다축 맥락 복원을 skill로 호출한다. session-recap, NEXT.md·ROADMAP.md, git status/log, semantic-memory를 조합해 어제/오늘의 작업축을 되살린다.

6SKILL.mdUpdated May 28, 2026

junghan0611/forge

tools

VerifiedTrustedCommunity

포지(forge) — 셀프호스팅 Forgejo 이슈/PR/라벨/코멘트 작업면. 힣 에이전트가 코드 면에서 일하는 공유 손. botment의 자식 패턴(가든 댓글면 → 코드 댓글면 endpoint swap). 'forge', 'forgejo', '포지', 'issue', 'pull request', '이슈', '풀리퀘', '코드 댓글', '라벨'.

6SKILL.mdUpdated May 27, 2026

junghan0611/entwurf-peek

tools

VerifiedTrustedCommunity

sync entwurf 자식을 들여다보는 손. 호출자가 'Mattering...'에 묶여있을 때 자식 분신이 무엇을 하는지 확인. entwurf_peers는 control socket 있는 세션만 보여주는데, 이 스킬은 자식 entwurf 세션(이름 태그 entwurf)까지 합쳐서 본다. 트리거: 'entwurf-peek', '분신 들여다보기', '진행 중 세션', 'sync entwurf 안에서', 'peek session', 'live session map', '분신 추적'.

6SKILL.mdUpdated May 1, 2026

junghan0611/entwurf-peek

Download

For Claude Desktop. Download once, then upload the file in the app — no terminal needed.

Need help? View full Cowork setup guide →

Install manually

Choose your platform

# Clone the repo
git clone https://github.com/junghan0611/agent-config.git

# Copy into Claude Code skills folder (global)
cp -r agent-config/skills/commit ~/.claude/skills/

Claude Code Skills — official skills path docs.

Repository

junghan0611/agent-config

6 stars

Compatible with

Claude Code

OpenAI Codex CLI

ChatGPT