Directory
Skills and plugins
Search the directory and filter by grade or ecosystem. Live entries are A, B, C and D. No F grades are published yet. 15739 entries are listed. Only entries the scanner has graded are listed.
Grades are a hygiene check, not a security guarantee.
This grade is a hygiene screen, not proof of safety.
Showing 901 to 950 of 1,212 Claude Code, Grade C entries. Page 19 of 25.
Listed A to Z by name.
No entries match that filter.
Grade C, One medium finding, Hexagon
NETWORK_HOST_FROM_INPUT
scripts/verify-registry.mjs:5 const response = await fetch(`https://registry.npmjs.org/${pkg.name}/${pkg.versiroutinekit (Skill)
Compatibility: Not stated.
**Your agent did it once.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED
scoville-plan/scripts/check_text_size.py:102 result = subprocess.run(command, stdout=subprocess.PIPE, stderr=sScoville Plan (Skill)
Compatibility: Not stated.
Scoville Plan keeps goals, decisions and progress in the repository so longer work survives the next conversation.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
scripts/sync-rules.mjs:13 const res = await fetch(url, { headers: { "user-agent": "secretgate-sync-rules" } })secretgate (Skill)
Compatibility: Not stated.
**Local secrets firewall for coding agents.** Detects credentials in prompts, file reads and tool output **before they are sent to the LLM API**, redacts them to stable placeholders, and restores the real values locally when the agent writes them back.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST
skills/model-connector/scripts/fast_connect.py:68 os.remove(tmp)Sichen Skills (Skill)
Compatibility: Not stated.
一组经过真实项目验证的 Claude Code / Agent Skills（SKILL.md 开放标准，兼容 Claude Code、Codex CLI、Cursor、OpenClaw 等 20+ Agent）。
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED
scripts/demo_change_review.py:27 subprocess.run(["git", "-C", str(root), *command], check=True, capture_outputSkill Doctor (Skill)
Compatibility: Not stated.
**The CI quality gate for Agent Skills.**
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV
agent_skill_radar/github.py:136 result = subprocess.run(Skill Radar (Skill)
Compatibility: Not stated.
Use when the user wants to scan GitHub for AI agent skills, MCP servers, prompt systems, context engineering tools, or agent-native CLI opportunities; generate daily opportunity reports; choose a two-week open-source project; or turn GitHub trend research into build briefs, fake-
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED SECRETS_PATH_DOTENV
internal/updater/updater.go:72 cmd := exec.Command("/bin/sh", "-c", command)Skill Router (Skill)
Compatibility: Not stated.
ALWAYS LOADED.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST SECRETS_PATH_CREDENTIAL
SKILL.md:98 curl -s "https://lightmake.site/api/v1/search?q=<关键词>&limit=10"Skill Scout (Skill)
Compatibility: Not stated.
上下文感知的 Skill 推荐器。读取当前 Agent 身份与项目处境，理解用户卡点，去 GitHub 和公开注册表检索可用 Skill，评分排序后推荐 Top 3 并说明能解决到什么程度。当用户说"帮我找个 skill"、"装个能做 X 的技能"、"有没有解决 X 的 skill"、"UI 好丑有没有 skill"、"这种情况该装什么 skill"，或描述一个反复卡住的痛点并希望有现成能力包时使用。只推荐，不自动安装。
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST +1 more
tools/benchmark_gitskills.py:421 shutil.rmtree(workspace, ignore_errors=True)skilllineage (Skill)
Compatibility: Not stated.
**Trace the copies, variants, and lineage evidence of AI Agent Skills.**
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_REMOTE_SCRIPT
plugins/agy/evals/bench/cloud-setup.sh:12 curl -fsSL https://antigravity.google/cli/install.sh | bash -s -- --skills (Skill)
Compatibility: Not stated.
Standalone Claude Code skills: grounded read-back before edits, a Tailwind v4 house style, a one-brief handoff across /clear, and a measured cut of what every request carries.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST
skills/carrier-scorecard/scripts/test_carrier_scorecard.py:59 result = subprocess.run([sys.executable, str(HERskills (Skill)
Compatibility: Not stated.
The skill catalog for **Aria Code** , Arthera's AI quant terminal.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED
skills/image-to-svg/scripts/convert_image_to_svg.py:191 subprocess.run(command, check=True, stdout=subprocess.skills (Skill)
Compatibility: Not stated.
[English](README.en.md) | 简体中文
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST +1 more
skills/devto-post/scripts/make-cover.py:97 subprocess.run(skills (Skill)
Compatibility: Not stated.
Claude Code skills I use day to day, packaged as a plugin marketplace so each one can be installed on its own.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST +3 more
_quarantine/detached-child/SKILL.md:46 subprocess.Popen([sys.executable, _DETACHED], cwd=root,skills (Skill)
Compatibility: Not stated.
<p> <picture> <source media="(prefers-color-scheme: dark)" srcset="assets/banner-dark.svg"> <img alt="skills.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST SECRETS_PATH_CREDENTIAL +1 more
scripts/npm-package/bin/cli.js:28 https.get({ hostname: opts.hostname, path: opts.pathname + opts.search, headskills (Skill)
Compatibility: Not stated.
An open-source library of AI agent skills following the [Agent Skills](https://agentskills.io) open standard.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST +1 more
chrome-cdp-setup/scripts/setup_chrome_cdp.sh:29 if subprocess.run(["pgrep", "-x", "Google Chrome"], capture_ouskills (Skill)
Compatibility: Not stated.
Skills shared by soulmachine for improving daily work efficiency with AI Agents (Claude Code, Codex, etc.).
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED NETWORK_NO_FIXED_HOST
scripts/test_release.py:24 result = subprocess.run(Socialbu Agent (Skill)
Compatibility: Not stated.
Manage authorized SocialBu posts, accounts, queues, analytics, automations, curated content, media, and social listening.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_HOST_FROM_INPUT
agents/web-claim-verifier.agent.md:69 {"kind": "tool", "tool": "bash", "input_head": "<optional short target -Solidforge Pi (Skill)
Compatibility: Not stated.
**SolidForge for the [Pi](https://github.com/earendil-works/pi-mono) coding agent** , an independent, pi-native implementation of the SolidForge Loop Engineering system: the **converge → specify → implement** pipeline plus two additive **outcome-axis** layers (cited-source verif
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK SECRETS_PATH_DOTENV
package.json:17 preparestem (Skill)
Compatibility: Not stated.
Engineering discipline for coding agents: size the work, keep every change on the record, verify against intent, never lose a session.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_DELETE_FILES INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST +1 more
packages/core/src/parsers/codex.test.ts:719 fs.rmSync(file);tokmeter (Skill)
Compatibility: Not stated.
>-
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED NETWORK_NO_FIXED_HOST SECRETS_PATH_UNCONFIRMED
skills/tudouni-lens/scripts/onboarding.py:317 process = subprocess.Popen([sys.executable, str(Path(__file__).rTudouni Lens (Skill)
Compatibility: Not stated.
土豆泥图视：在对话中开启图片与视频创作，让每次灵感都有清晰的下一步。
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV
package.json:17 postinstallTwist Cli (Skill)
Compatibility: Not stated.
<p align="center"> <img src="icons/twist-cli.png" alt="Twist CLI" width="150" height="150" /> </p>
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_REMOTE_SCRIPT
scripts/fixtures/lane-breaches/scripts/net-herestring-shell.sh:6 cat <<< 'curl https://example.invalid' | bashUncle Bob Skills (Skill)
Compatibility: Not stated.
Fifty evidence-forged skills for human-directed agent engineering
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST SECRETS_PATH_UNCONFIRMED +1 more
CLAUDE.md:192 /runner.sh → GET/HEAD the reference cron+curl runner (text/plain)vellum (Skill)
Compatibility: Not stated.
A shared space where AI agents leave short thoughts, weave lineages from each other's words, and can come back to see what became of what they said.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
scripts/ingest_vk.py:87 def fetch(section: str) -> list[dict]:Vk Mcp Ru (Skill)
Compatibility: Not stated.
VK API прямо в Claude: товары магазина сообщества, посты, рекламные кампании и статистика, диалоги с клиентами, лид-формы.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST
skills/whetstone-mcp/SKILL.md:362 auth = subprocess.run(["claude", "auth", "status"], env=probe_env,Whetstone Mcp (Skill)
Compatibility: Not stated.
Evaluate an MCP server across a weak and a strong model, then close the gap by changing the server.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES SECRETS_PATH_DOTENV
skills/computer-history-collector/scripts/computer_history_collector.py:207 result = subprocess.run(A Particular Set Of Skills (Skill)
Compatibility: Not stated.
Models and agents can be replaced.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV
package.json:55 prepareadversary (Skill)
Compatibility: Not stated.
Adversarial QA for AI-generated software , the QA engineer your coding agent can't fool.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST +1 more
evals/_harness/run_skill.py:75 proc = subprocess.run(Advertising Skills (Skill)
Compatibility: Not stated.
Skills for paid advertising , channel selection, budget policy, audience design, creative development, and measurement across search, paid social, video, native, and AI-answer surfaces.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
install.sh:31 curl -fsSL "${release_base}/${asset}" -o "${archive}"Agent Preview (Skill)
Compatibility: Not stated.
Automatically return a temporary browser link after Claude Code creates an HTML page, static site, or other browser-previewable artifact.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_HOST_FROM_INPUT NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST +1 more
skills/implementing-legend-state/references/observables.md:136 const server$ = observable(() => fetch('url').tAgent Skills (Skill)
Compatibility: Not stated.
Opinionated [agent skills](https://agentskills.io) for Claude Code, Cursor, Codex, OpenCode, and any other agent that speaks the [skills.sh](https://skills.sh) format.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_DELETE_FILES INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST +1 more
scripts/check-installs.py:22 return subprocess.run(['git', *args], stdout=subprocess.PIPE, text=True, check=TrAgent Skills (Skill)
Compatibility: Not stated.
Personal agent skills for planning, building, reviewing, and explaining software.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_REMOTE_SCRIPT
row-bot-docker-setup/README.md:218 curl https://ollama.ai/install.sh | shAgent Skills (Skill)
Compatibility: Not stated.
A personal, curated library of skills for AI agents.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED SECRETS_PATH_UNCONFIRMED
skills/explainer/scripts/align_narration.py:32 duration = float(subprocess.check_output(["ffprobe", "-v", "errAgent Skills (Skill)
Compatibility: Not stated.
Portable skill library for architecture, code quality, prompt maintenance, research, interview-led builds, and multi-agent workflows.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST +2 more
.agents/hooks/markdown_lint_check.py:173 result = subprocess.run(Agent Standards (Skill)
Compatibility: Not stated.
<div align="center">
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_REMOTE_SCRIPT
.github/workflows/validate.yml:18 curl -fsSL https://claude.ai/install.sh | bashAgent Toolkit (Skill)
Compatibility: Not stated.
Personal agent toolkit: individually installable skills for Claude Code and Codex.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_HOST_FROM_INPUT
apps/cli/skills/ip-lookup/scripts/invoke.js:26 const res = await fetch(url, { headers: { 'User-Agent': 'AgentOAgentOctopus (Skill)
Compatibility: Not stated.
> Intelligent routing layer that connects user needs to Skills , install once, works everywhere.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST
README.md:27 curl -fsSL https://github.com/storious/agulater/releases/download/v0.2.1/install.sh | shagulater (Skill)
Compatibility: Not stated.
Agulater is an optional lifecycle and package tool for [Agul](https://github.com/storious/agul).
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV OBFUSCATION_MINIFIED_PAYLOAD
src/WatermarkRemover.CLI/wwwroot/_astro/format.DyQ9F1g-.js:1 Type","application/json");const a=await fetch(`${Ai Watermark Remover (Skill)
Compatibility: Not stated.
<div align="center">
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV OBFUSCATION_MINIFIED_PAYLOAD
skills/plan-agent-tasks/assets/dashboard/resources/panel.js:267 const response=await fetch('api/status',{cacheAi Workflow Skills (Skill)
Compatibility: Not stated.
Reusable workflows for AI agents, with task cards, examples, and explicit execution boundaries.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV
docs/backends.md:39 's global bin. The official installer (`curl -fsSL https://x.ai/cli/install.sh | bash`) inaibridge (Skill)
Compatibility: Not stated.
<div align="center"> <img src="assets/logo.png" alt="aibridge logo" width="140" />
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED INSTALL_UNPINNED
scripts/retouch.py:104 subprocess.Popen(['open', path])Aie Skills (Skill)
Compatibility: Not stated.
专业修图师技能。支持查看直方图、提取EXIF镜头信息、调整色温/高光/阴影/曝光/对比度/饱和度、局部修图（径向/矩形局部调整、瑕疵修复、磨皮）、降噪、丁达尔光（体积光）、构图裁剪、青橙/冷灰等颜色分级、观察图像，并支持RAW文件。
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
references/02-testflight-ios.md:80 Watch and Live Activity widgets need their own provisioning profiles in `crApp Store Review Playbook (Skill)
Compatibility: Not stated.
End-to-end guide for building and shipping a mobile app to the App Store and Google Play, written from real rejections and real build failures.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV
00_meta/scripts/ingest_github_repo.py:61 # Using subprocess.run for better control and error handlingarchivum (Skill)
Compatibility: Not stated.
<p align="center"> <a href="https://antreas.io/archivum/"> <img src="assets/archivum-mimic.png" width="160" alt="Archivum’s winged archive mimic mascot"> </a> </p>
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED SECRETS_PATH_DOTENV
tools/verify_public_repo.py:56 result=subprocess.run(Asic Intelligence Plugin Skill Package (Skill)
Compatibility: Not stated.
**Powered by ASIC Intelligence** **Read-only Bitcoin ASIC diagnostics from the evidence you already have.**
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
SECRETS_PATH_DOTENV OBFUSCATION_MINIFIED_PAYLOAD
.gitignore:7 .envAudit Reasoning (Skill)
Compatibility: Not stated.
Audit a local Codex session or response turn using trace timing events, reporting activity percentages and locating unclassified gaps between model output, tool calls, tool results and turn boundaries.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
INSTALL_LIFECYCLE_HOOK
package.json:39 prepareazdospec (Skill)
Compatibility: Not stated.
Remote-first spec-driven development for Azure DevOps.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV OBFUSCATION_ENCODED_BLOB
packages/cli/src/v2-commands.ts:74 const response = await dependencies.fetch(input, { ...init, signal: controlbbbbb (Skill)
Compatibility: Not stated.
[English](README.md) | [简体中文](README.zh-CN.md) | [Español](README.es.md) | [日本語](README.ja.md) | [Deutsch](README.de.md) | [Français](README.fr.md) | [Português (Brasil)](README.pt-BR.md)
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
NETWORK_NO_FIXED_HOST SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
bin/kaiva-bridge.js:25 const bridgeFetch = (url, init = {}) => fetch(url, { ...init, headers: { 'User-Agent':bridge (Skill)
Compatibility: Not stated.
Give Claude Code real access to an API, a database, documents or a remote MCP server through Kaiva Bridge hosted MCP servers.
This grade is a hygiene screen, not proof of safety.
Grade C, One medium finding, Hexagon
PERMISSION_SHELL_UNDECLARED PERMISSION_DELETE_FILES NETWORK_NO_FIXED_HOST +1 more
superqa_tui/cli.py:211 subprocess.run(["open" if sys.platform == "darwin" else "xdg-open", path],Browser Qa (Skill)
Compatibility: Not stated.
Browser QA with reviewable YAML scenarios.
This grade is a hygiene screen, not proof of safety.
