Skip to content

hgxszhj/https-github.com-zhaoxuya520-reverse-skill / skills

reverse-skill-router

Flagged

Routes reverse engineering, exploitation, penetration testing, malware, mobile, firmware, browser automation, documentation, and security tasks to the appropriate specialist skill. Use when a task spans modules or the correct reverse-skill entrypoint is unclear.

FileKindSize
CONTRIBUTING.mdmarkdown20.4 KB
INDEX.mdmarkdown11.2 KB
MASTER-ROUTING.mdmarkdown7.3 KB
SKILL.mdskill md12.9 KB
config/routing.jsonconfig14.7 KB
field-journal/2026-05-15-cellular-pro-mumu-ksad-fragment-fix.mdmarkdown1.3 KB
field-journal/2026-05-15_lumine-go-reverse.mdmarkdown2.2 KB
field-journal/2026-05-16_pentest-personalblog-fun-mass-assignment.mdmarkdown5.1 KB
field-journal/2026-05-17_pentest-vue-spa-actuator-leak.mdmarkdown3.0 KB
field-journal/2026-05-25_pentest-cf-access-sibling-subdomain-cookie-poisoning.mdmarkdown12.0 KB
field-journal/2026-05-26_pentest-newapi-rate-limit-bypass.mdmarkdown2.9 KB
field-journal/2026-06-29_burp-mcp-full-test-and-fix.mdmarkdown7.4 KB
field-journal/2026-07-05_dsl-vm-captcha-reverse.mdmarkdown3.5 KB
field-journal/2026-07-14_android-arm64-self-extract-source-recovery.mdmarkdown5.7 KB
field-journal/2026-07-18_gin-juice-client-friction.mdmarkdown1.6 KB
field-journal/2026-07-20_reverse-toolchain-windows-bootstrap.mdmarkdown3.9 KB
field-journal/2026-07-22_electron-bytenode-privileged-update-chain.mdmarkdown6.7 KB
field-journal/2026-08-01_nextjs-cdk-contract-faithful-local-reconstruction.mdmarkdown7.6 KB
field-journal/2026-08-01_pentest-encryption-oracle-public-template-sql-admin-takeover.mdmarkdown8.1 KB
field-journal/2026-08-06_cortex-m-msc-firmware-self-keyed-rotate-xor.mdmarkdown4.8 KB
field-journal/2026-08-08_client-neutral-structured-routing-pr-integration.mdmarkdown4.1 KB
field-journal/2026-08-08_pr-value-triage-and-safe-integration.mdmarkdown1.7 KB
field-journal/2026-08-14_windows-powershell-native-exit-code-pr-review.mdmarkdown5.5 KB
field-journal/2026-08-17_tool-index-r2-fallback-powershell-pwsh-host-fix.mdmarkdown12.1 KB
field-journal/2026-08-20_apk-flutter-banner-ad-removal.mdmarkdown6.0 KB
field-journal/2026-09-03_issue-pr-security-boundary-integration.mdmarkdown2.9 KB
field-journal/CONTRIBUTE-BACK.mdmarkdown5.3 KB
field-journal/_index.mdmarkdown6.1 KB
field-journal/_template.mdmarkdown4.4 KB
field-journal/anonymization.mdmarkdown6.6 KB
field-journal/precedent-auth.mdmarkdown1.7 KB
field-journal/precedent-pentest.mdmarkdown25.0 KB
field-journal/precedent-reverse.mdmarkdown13.2 KB
field-journal/seed-001_elf-packed-loader.mdmarkdown3.3 KB
field-journal/seed-002_go-malware-stripped.mdmarkdown3.4 KB
field-journal/seed-003_web-api-auth-bypass.mdmarkdown3.2 KB
field-journal/seed-004_js-sign-webpack.mdmarkdown3.6 KB
field-journal/seed-005_ad-certipy-esc1.mdmarkdown2.0 KB
field-journal/seed-006_ssrf-cloud-metadata.mdmarkdown2.7 KB
field-journal/seed-007_ntlm-relay-coercer.mdmarkdown2.9 KB
field-journal/seed-008_apk-okhttp-ssl-pin-bypass.mdmarkdown4.4 KB
field-journal/seed-009_ios-jailbreak-detect-bypass.mdmarkdown4.6 KB
field-journal/seed-010_ctf-pwn-rop-x64.mdmarkdown5.0 KB
field-journal/seed-011_pcap-protocol-reverse.mdmarkdown4.5 KB
field-journal/seed-012_log4shell-jndi-rce.mdmarkdown5.1 KB
field-journal/seed-013_kerberoasting-spn.mdmarkdown4.6 KB
field-journal/seed-014_unity-il2cpp-reverse.mdmarkdown5.3 KB
field-journal/seed-015_iot-firmware-uart.mdmarkdown6.0 KB
field-journal/seed-016_k8s-container-escape.mdmarkdown6.6 KB
field-journal/seed-017_xxe-oob-exfil.mdmarkdown5.3 KB
ops/IDENTITY.mdmarkdown2.7 KB
ops/README.mdmarkdown1.3 KB
ops/analysis-blindspot-cookbook.mdmarkdown4.1 KB
ops/analysis-decision-framework.mdmarkdown5.4 KB
ops/evidence-finding-path.mdmarkdown4.3 KB
ops/role-map.mdmarkdown3.1 KB
ops/sandbox-profile.mdmarkdown2.1 KB
ops/scope-contract.mdmarkdown3.9 KB
ops/skill-supply-chain.mdmarkdown3.6 KB
ops/timeline-workitem.mdmarkdown3.5 KB
references/community-security-skills.mdmarkdown7.5 KB
references/domain-coverage-map.mdmarkdown2.9 KB
routing.mdmarkdown27.1 KB
routing_zh.mdmarkdown29.5 KB
scripts/append-evidence.ps1script7.1 KB
scripts/bootstrap-manifest.jsonconfig12.5 KB
scripts/bootstrap-reverse.ps1script41.0 KB
scripts/bootstrap-reverse.shscript34.1 KB
scripts/case-guard.ps1script3.9 KB
scripts/case-guard.shexecutablescript3.5 KB
scripts/case-init.ps1script14.4 KB
scripts/case-init.shexecutablescript12.3 KB
scripts/consolidate-evidence.ps1script2.7 KB
scripts/consolidate_evidence.pyscript3.1 KB
scripts/extract-summaries.ps1script5.5 KB
scripts/lib/BootstrapSupplyChain.ps1script7.1 KB
scripts/lib/HostRuntime.ps1script1.9 KB
scripts/lib/RouteScope.ps1script718 B
scripts/lib/ToolDiscovery.ps1script44.6 KB
scripts/lib/WorkRoot.ps1script876 B
scripts/master-route.ps1script7.8 KB
scripts/master-route.shexecutablescript5.4 KB
scripts/refresh-tool-index.ps1script7.7 KB
scripts/refresh-tool-index.shscript17.2 KB
scripts/scan-leaks.ps1script2.9 KB
scripts/smoke.ps1script8.9 KB
scripts/test-bash-workflow.shexecutablescript2.7 KB
scripts/test-bootstrap-codex-encoding.ps1script5.0 KB
scripts/test-bootstrap-manifest.shscript12.5 KB
scripts/test-bootstrap-supply-chain.ps1script10.9 KB
scripts/test-client-neutral-bootstrap.ps1script3.5 KB
scripts/test-client-neutral-bootstrap.shscript9.5 KB
scripts/test-consolidate-evidence.ps1script3.1 KB
scripts/test-p0-friction.ps1script22.4 KB
scripts/test-parse-contracts.ps1script7.6 KB
scripts/test-routing.ps1script3.8 KB
scripts/test-routing.shscript4.8 KB
scripts/test-workflow-title-safety.ps1script1.8 KB
scripts/update-star-history.ps1script5.4 KB
scripts/verify-doc-facts.ps1script4.2 KB
scripts/verify-doc-links.pyscript3.6 KB
scripts/verify-repository-security.pyscript7.9 KB
scripts/verify-routing-coherence.ps1script33.7 KB
tests/routing-benchmark.jsonconfig17.3 KB
tool-index.md.templateother2.8 KB

references/community-security-skills.md

View on GitHub
# 社区安全 Skill 生态对照(2026-07)

> 来源检索日期:**2026-07-17**  
> 目的:让 reverse-skill **知道外面有什么**,按需借鉴,**不**把外部巨型库整仓并入本包。  
> 本包身份:路由 + 工具自举 + 证据/scope 契约 + field-journal(见 `ops/IDENTITY.md`)。

## 1. 外部高价值仓库(可学习,勿盲装)

| 仓库 | 规模/定位 | 对本包价值 | 风险 |
|------|-----------|------------|------|
| [trailofbits/skills](https://github.com/trailofbits/skills) | ToB 安全研究 Claude 插件市场 | 审计/漏洞分析/RE 插件质量标杆 | 需按 ToB 市场安装;勿默认信任非 curated 副本 |
| [trailofbits/skills-curated](https://github.com/trailofbits/skills-curated) | 已审核插件列表 | 优先于任意社区 skill | 同上 |
| [Orizon-eu/claude-code-pentest](https://github.com/Orizon-eu/claude-code-pentest) | 6 个 pentest 生命周期 skill + 纯 Python 脚本 | 侦察→利用→报告流水线可对标我们 `attack-chain`+`pentest-tools` | 授权边界要自检;脚本需沙箱 |
| [trilwu/secskills](https://github.com/trilwu/secskills) | 16 skills + 6 专家 subagent | 多角色分工可对标 `ops/role-map.md` | 插件形态,与本包 monorepo 不同 |
| [Masriyan/Claude-Code-CyberSecurity-Skill](https://github.com/Masriyan/Claude-Code-CyberSecurity-Skill) | ~15–19 领域 skill(含 RE/OT/CSOC) | 领域覆盖 checklist | 深度不如本包单域 skill |
| [mukul975/Anthropic-Cybersecurity-Skills](https://github.com/mukul975/Anthropic-Cybersecurity-Skills) | **800+** skill · ATT&CK/NIST 映射 | **框架映射**与领域目录可参考,不宜整库依赖 | 体量过大、维护与投毒面巨大 |
| [Eyadkelleh/awesome-skills-security](https://github.com/Eyadkelleh/awesome-claude-skills-security) | SecLists 打包为 agent skills | 字典/payload 入口 | 与 seclists bootstrap 重叠 |
| [securityfortech/awesome-security-skills](https://github.com/securityfortech/awesome-security-skills) | 安全 skill 精选列表 | 发现新 skill 的索引 | 列表型,需逐个审计 |
| [VoltAgent/awesome-agent-skills](https://github.com/VoltAgent/awesome-agent-skills) | 1000+ 跨厂商 skill 索引 | 发现官方/社区 skill | 非安全专用 |
| [anthropics/claude-code-security-review](https://github.com/anthropics/claude-code-security-review) | PR 安全审查 GitHub Action | 可对标我们 docs/报告侧「变更审计」场景 | CI 产品,不是 RE 路由 |
| [agentskills.io](https://agentskills.io) | Agent Skills 开放标准 | frontmatter/目录约定对齐 | 标准本身无攻防内容 |

### 1.1 第二轮检索增补(2026-07-17 再搜)

| 仓库 / 资源 | 定位 | 本包落点 |
|-------------|------|----------|
| [trailofbits/skills](https://github.com/trailofbits/skills) 插件:`audit-context-building` `differential-review` `semgrep-rule-creator` `sharp-edges` `dwarf-expert` `burpsuite-project-parser` | 审计上下文、差分安全审查、危险 API、DWARF、Burp 工程解析 | 对照 `ida-reverse`/`docs-generator`/审计工作流;**不**整库并入 |
| [HexRaysSA/ida-claude-code-plugins](https://github.com/HexRaysSA/ida-claude-code-plugins) | 官方 IDA Claude 插件(含 domain 自动化,标注 unsafe) | `ida-reverse` MCP 路径对照;unsafe 插件默认不启用 |
| [P4nda0s/reverse-skills](https://github.com/P4nda0s/reverse-skills) | IDA-NO-MCP:导出反编译后再分析;rev-frida/dex-dump/u3d | 与「MCP 不可用时的离线导出」互补 |
| [2389-research/binary-re](https://github.com/2389-research/binary-re) | triage→static(r2/Ghidra)→dynamic(QEMU/GDB/Frida)→synthesis | `reverse-engineering` 阶段门闩见 `re-agent-workflow.md` |
| [incogbyte/android-reverse-engineering-claude-skill](https://github.com/incogbyte/android-reverse-engineering-claude-skill) | APK 解包、端点提取、自适应 Frida 绕过 | 对照 `apk-reverse`;动态脚本需 scope |
| [OwenPawl/cerberus-re-skill](https://github.com/OwenPawl/cerberus-re-skill) | Apple 向 Ghidra+LLDB+Frida 三循环 | 可参考 macOS/iOS 动态环 |
| [ljagiello/ctf-skills](https://github.com/ljagiello/ctf-skills) | CTF reverse/pwn;工具按需装 | 对照 CTF-Sandbox + `pwn-chain` |
| [shuvonsec/claude-bug-bounty](https://github.com/shuvonsec/claude-bug-bounty) | /recon→/hunt→/validate→/report | 对照 `recon-pipeline.md` + scope 门 |
| [PayloadsAllTheThings](https://github.com/swisskyrepo/PayloadsAllTheThings) | Web payload + Prompt Injection 章节 | `pentest-tools/payloads` 优先;LLM 见 `llm-security` |
| [HackTricks](https://hacktricks.wiki/) | 渗透方法论 + **AI/MCP 滥用** | 见 skill-supply-chain MCP 节 |
| [appsecsanta AI pentesting agents 2026](https://appsecsanta.com/research/ai-pentesting-agents-2026) | 39+ 开源 AI 渗透 agent 架构分类 | 多 agent ≠ 必须;我们用 role-map |
| Snyk 评测「更多 skill ≠ 更好」 | 技能堆叠可能降低审计质量 | 强化「深 skill + 路由」策略 |

## 2. 安全标准与威胁(2025–2026)

| 来源 | 要点 | 本包落点 |
|------|------|----------|
| [OWASP Agentic Skills Top 10](https://owasp.org/www-project-agentic-skills-top-10/) | 恶意 skill、供应链、权限滥用、记忆投毒等 | `ops/skill-supply-chain.md` |
| [Anthropic Agent Skills 工程文](https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills) | 只装可信源;审脚本与依赖 | 同上 + bootstrap 禁止猜路径 |
| ClawHavoc 等投毒活动(AST10 记载) | 注册表批量恶意 skill | 禁止从未知 registry 一键安装进本包 |

## 3. 本包已有 vs 外部「广」覆盖

| 领域 | reverse-skill | 外部常有、我们不整库并入的原因 |
|------|---------------|--------------------------------|
| APK/JS/IDA/r2/固件/pwn | **深** skill + 脚本 | 保持深度与 tool-index 绑定 |
| 渗透/攻击链/SRC | pentest-tools + attack-chain + src-hunter | Orizon 类可作方法论对照 |
| LLM/Agent 安全 | llm-security | AST10 增强 skill 自身安全 |
| 证据/scope/角色 | **ops/**(特色) | 多数 skill 包没有 case 契约 |
| OT/ICS / 纯 GRC / 欺诈 F3 | 无独立 skill | 路由未命中 → 提议新增或外链,不硬塞 |
| 800+ 微 skill | 不复制 | 用 MASTER 路由 + 域 skill 替代碎片化 |

## 4. 借鉴规则(MUST)

```text
1. 禁止 git submodule 拉整库 800+ skills 当运行时依赖
2. 借鉴时:提取「阶段/清单/命令模式」写入本包 references 或现有 skill
3. 外部脚本:先在隔离环境看依赖与网络行为,再考虑 bootstrap-manifest
4. 新场景:走 CONTRIBUTING 新增 skill,并更新 routing + RULES 关键词
5. 标注来源 URL + 检索日期(本文件格式)
6. 安装/合并前走 ops/skill-supply-chain.md 清单
7. 运行时只加载 MASTER-ROUTING 的 PRIMARY(+必要 secondary),避免技能堆叠过载
```

## 4.1 本包已沉淀的「借鉴产物」(非外库依赖)

| 产物 | 路径 |
|------|------|
| RE 四阶段 | `reverse-engineering/references/re-agent-workflow.md` |
| 授权侦察 | `pentest-tools/references/recon-pipeline.md` |
| 攻击链门闩 | `attack-chain/references/lifecycle-checklist.md` |
| Skill 供应链 | `ops/skill-supply-chain.md` |
| 领域覆盖 | `references/domain-coverage-map.md` |

## 5. 建议优先级(后续迭代)

| 优先级 | 动作 |
|--------|------|
| P0 已做 | ops 契约、MASTER 路由、skill 供应链安全文档 |
| P1 | 对照 Orizon/ToB 补 pentest 阶段检查单到 attack-chain references |
| P2 | 可选「外链 skill 白名单」配置,不进默认路径 |
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall