Menu

Not in the top 1,000 by GitHub stars

0rangec3t/Black-cat

Claude Code RedTeam Skill — Hypothesis-Driven Cognitive Architecture。一个假设-证据驱动的红队skill

Claude CodeSecurityEngineering

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

skills/pentest-redteam/SKILL.md

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

80392399070a → 80392399070a

Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 80392399070a

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit 80392399070a0241aaeb6394e39f32c0843906bc

git clone --filter=blob:none --no-checkout https://github.com/0rangec3t/Black-cat.git .hotskills/0rangec3t-black-cat && git -C .hotskills/0rangec3t-black-cat checkout --detach 80392399070a0241aaeb6394e39f32c0843906bc # Copy each selected skill directory from .hotskills/0rangec3t-black-cat into .claude/skills/<skill-name>/
328GitHub stars
0stars in 7 days
1skill in this repository
0commits in 30 days

Heat

0.4
  • Star growth, last 7 daysNot available
  • Star growth, last 24 hoursNot available
  • Rank movement, last 7 daysNot available
  • Commits, last 30 daysNot available
  • Accounts asking about it through MCP, last 7 daysNot available
  • Hot votes this weekNot available

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Oct 25, 2026Oct 31, 2026328

30-day change not yet available: not enough history.

Individual skills in this repository

How it works

pentest-redteam

When your agent uses it

授权红队假设驱动执行框架。按信号显式路由到一个 technique 文件,活跃目录默认 1 个、最多 2 个。

What it loads into context

Frontmatter

63 tokens

Always available to the agent

SKILL.md

2,424 tokens

When the skill is activated

Reference documents

27,483 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. 1. 授权 Gate（一次性）
  2. 2. 模式选择
  3. 3. 显式路由
  4. 4. 上下文约束
  5. 5. 运行时适配与工具降级
  6. 硬约束（跨模式，始终生效）
  7. L2：Attack State Machine（本 skill 的核心设计）

Scripts executed separately

  • scripts/case_ledger.py

Files

  • SKILL.mdcore8,779 bytes
  • SKILL.mdcore8,779 bytes
  • scripts/case_ledger.pyruns24,056 bytes
  • techniques/ad.mddoc18,279 bytes
  • techniques/cloud.mddoc8,060 bytes
  • techniques/database.mddoc7,424 bytes
  • techniques/evasion.mddoc20,675 bytes
  • techniques/recon.mddoc13,949 bytes
  • techniques/reversing.mddoc6,145 bytes
  • techniques/web.mddoc12,740 bytes
  • templates/engagement-report.mddoc283 bytes
  • templates/finding-report.mddoc512 bytes

No excerpt is shown for this license. Read the source on GitHub.

View the complete source on GitHub

Skills included

  • pentest-redteam

Repository facts

License
No license detected
Created
Jul 30, 2026
Last pushed
Aug 3, 2026
Forks
43
Open issues
1
Project activity
Slowing

Before you install

Network or broad shell

Automated signals, not a guarantee

Offensive security tool: run only in authorized labs.

Checked Nov 4, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 1 note
  • skills/pentest-redteam/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 57 signals by file

skills/pentest-redteam/SKILL.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash,WebFetch,WebSearch

skills/pentest-redteam/scripts/case_ledger.py

  • Executable scriptline 1
    #!/usr/bin/env python3

skills/pentest-redteam/techniques/ad.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash,WebFetch
  • Network accessline 16
    2. 收集 > 移动 > 提权 —— 先看清全局（BloodHound），不要拿到 shell 就立刻提权 3. OPSEC 不是"小声点"，是"看起来正常"——内存操作、原生 API 调用、LDAP 流量才是正常运维行为 4. 参考：https://lolol.farm/ — LOLBins/LOLDrivers/LOLRMM/LOT Tunnels/LOLEXFIL 综合索引。BOF 集合：TrustedSec CS-Situational-Awareness-BOF、a
  • Network accessline 18
    4. 参考：https://lolol.farm/ — LOLBins/LOLDrivers/LOLRMM/LOT Tunnels/LOLEXFIL 综合索引。BOF 集合：TrustedSec CS-Situational-Awareness-BOF、ajpc500/BOFs、outflanknl/C2-Tool-Collection
  • Network accessline 122
    - **信号**: 目标安装远程监控管理工具（TeamViewer/AnyDesk/ScreenConnect/MeshCentral 等） - **假设**: RMM 工具本身是合法远程管理通道，其流量和进程不受 EDR 审查 - **验证**: 参考 LOLRMM 项目 (https://lolol.farm/) → 枚举已安装 RMM → 提取 RMM 配置中的预共享密钥/agent 密钥 → 用 RMM 自带功能在目标间移动（它本来就是干这个的）
  • Network accessline 123
    - **假设**: RMM 工具本身是合法远程管理通道，其流量和进程不受 EDR 审查 - **验证**: 参考 LOLRMM 项目 (https://lolol.farm/) → 枚举已安装 RMM → 提取 RMM 配置中的预共享密钥/agent 密钥 → 用 RMM 自带功能在目标间移动（它本来就是干这个的） - **证实**: 通过 RMM 通道在目标间移动，EDR 零告警
  • Network accessline 124
    - **验证**: 参考 LOLRMM 项目 (https://lolol.farm/) → 枚举已安装 RMM → 提取 RMM 配置中的预共享密钥/agent 密钥 → 用 RMM 自带功能在目标间移动（它本来就是干这个的）
  • Network accessline 136
    - **信号**: 需要出网 C2，但出站规则严格 - **假设**: 可通过合法隧道服务建立 C2，流量看起来像 Microsoft/Cloudflare/Ngrok - **验证**: 参考 LOT Tunnels + LOLC2 项目 (https://lolol.farm/) → Dev Tunnels：`code --tunnel` 微软签名二进制 → WebSocket over SSH → 流量经 *.devtunnels.ms → Cloudflare Tun
  • Network accessline 137
    - **假设**: 可通过合法隧道服务建立 C2，流量看起来像 Microsoft/Cloudflare/Ngrok - **验证**: 参考 LOT Tunnels + LOLC2 项目 (https://lolol.farm/) → Dev Tunnels：`code --tunnel` 微软签名二进制 → WebSocket over SSH → 流量经 *.devtunnels.ms → Cloudflare Tunnel / Ngrok / Tailscale Fu
  • Network accessline 138
    - **验证**: 参考 LOT Tunnels + LOLC2 项目 (https://lolol.farm/) → Dev Tunnels：`code --tunnel` 微软签名二进制 → WebSocket over SSH → 流量经 *.devtunnels.ms → Cloudflare Tunnel / Ngrok / Tailscale Funnel 作为备选
  • Network accessline 212
    - **信号**: 目标使用 Entra ID（Azure AD）+ 本地 AD 混合 - **假设**: 云侧后门可绕过本地 AD 的安全加固 - **验证**: 参考 EntraGoat 六大场景 (https://github.com/Semperis/EntraGoat) → Service Principal 所有权滥用/App-Only Graph 权限提权/Group Membership 滥用/PIM 激活逃逸/Administrative Units 边界突
  • Network accessline 213
    - **假设**: 云侧后门可绕过本地 AD 的安全加固 - **验证**: 参考 EntraGoat 六大场景 (https://github.com/Semperis/EntraGoat) → Service Principal 所有权滥用/App-Only Graph 权限提权/Group Membership 滥用/PIM 激活逃逸/Administrative Units 边界突破/CBA 证书认证滥用 → Conditional Access 绕过（资源排除缺口
  • Network accessline 214
    - **验证**: 参考 EntraGoat 六大场景 (https://github.com/Semperis/EntraGoat) → Service Principal 所有权滥用/App-Only Graph 权限提权/Group Membership 滥用/PIM 激活逃逸/Administrative Units 边界突破/CBA 证书认证滥用 → Conditional Access 绕过（资源排除缺口 + sccauth 替代 Token Broker + A
  • Network accessline 227
    | 需求 | 参考项目 | 网址 | |------|---------|------| | WIndows 原生工具 | LOLBAS | https://lolbas-project.github.io/ |
  • Network accessline 228
    |------|---------|------| | WIndows 原生工具 | LOLBAS | https://lolbas-project.github.io/ | | AD 原生命令 | LOLAD | https://lolol.farm/ → LOLAD |
  • Network accessline 229
    | WIndows 原生工具 | LOLBAS | https://lolbas-project.github.io/ |
  • Network accessline 230
    | AD 原生命令 | LOLAD | https://lolol.farm/ → LOLAD |
  • Network accessline 231
    | 漏洞驱动 | LOLDrivers | https://www.loldrivers.io/ |
  • Network accessline 232
    | RMM 工具 | LOLRMM | https://lolol.farm/ → LOLRMM |
  • Network accessline 233
    | 隧道/C2 | LOT Tunnels + LOLC2 | https://lolol.farm/ → LOT Tunnels |
  • Network accessline 234
    | 数据外带方法 | LOLEXFIL | https://lolol.farm/ → LOLEXFIL |
  • Network accessline 235
    | Windows/AD 命令速查 | WADComs | https://wadcoms.github.io/ |
  • Network accessline 236
    | 持久化机制 | Persistence Info | https://lolol.farm/ → Persistence Info |
  • Network accessline 237
    | ESXi 原生工具 | LOLESXi | https://lolol.farm/ → LOLESXi |
  • Network accessline 238
    | Entra ID 攻击实验 | EntraGoat | https://github.com/Semperis/EntraGoat |

skills/pentest-redteam/techniques/cloud.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash,WebFetch
  • Network accessline 24
    - **信号**: 目标在 AWS/Azure/GCP 上运行（EC2/VM/GCE） - **假设**: IMDS 可访问且返回临时凭证 - **验证**: `curl http://169.254.169.254/latest/meta-data/iam/security-credentials/` (AWS) / `curl -H "Metadata:true" http://169.254.169.254/metadata/identity/oauth2/token`
  • Network accessline 25
    - **假设**: IMDS 可访问且返回临时凭证 - **验证**: `curl http://169.254.169.254/latest/meta-data/iam/security-credentials/` (AWS) / `curl -H "Metadata:true" http://169.254.169.254/metadata/identity/oauth2/token` (Azure) / `curl "http://metadata.google.int
  • Network accessline 26
    - **验证**: `curl http://169.254.169.254/latest/meta-data/iam/security-credentials/` (AWS) / `curl -H "Metadata:true" http://169.254.169.254/metadata/identity/oauth2/token` (Azure) / `curl "http://metadata.google.internal/computeMetadata/v1/i
  • Network accessline 38
    - **信号**: 发现任何域名或 JS 中引用云存储 URL - **假设**: 存在公开或弱权限的存储桶/容器 - **验证**: AWS S3 `aws s3 ls s3://bucket-name --no-sign-request` → Azure Blob `https://account.blob.core.windows.net/container` → GCP Storage `gsutil ls gs://bucket-name` → 阿里云 OSS `h
  • Network accessline 39
    - **假设**: 存在公开或弱权限的存储桶/容器 - **验证**: AWS S3 `aws s3 ls s3://bucket-name --no-sign-request` → Azure Blob `https://account.blob.core.windows.net/container` → GCP Storage `gsutil ls gs://bucket-name` → 阿里云 OSS `https://bucket.oss-<region>.aliyu
  • Network accessline 40
    - **验证**: AWS S3 `aws s3 ls s3://bucket-name --no-sign-request` → Azure Blob `https://account.blob.core.windows.net/container` → GCP Storage `gsutil ls gs://bucket-name` → 阿里云 OSS `https://bucket.oss-<region>.aliyuncs.com`

skills/pentest-redteam/techniques/database.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash
  • Dangerous shell operationline 65
    - **验证**: `CREATE EXTENSION plpythonu; CREATE FUNCTION exec(cmd text) RETURNS text AS $$ import subprocess; return subprocess.check_output(cmd, shell=True) $$ LANGUAGE plpythonu;` → 备选：Large Object 导出 + 编译 UDF 利用 libc 系统库（不引入外部文件）
  • Dangerous shell operationline 66
    - **证实**: exec('id') 返回命令输出

skills/pentest-redteam/techniques/evasion.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash

skills/pentest-redteam/techniques/recon.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash,WebFetch
  • Network accessline 18
    - **信号**: CNAME 指向 Cloudflare/Akamai/阿里云CDN/腾讯云CDN - **假设**: 源站 IP 隐藏但可通过证书/DNS 历史/特征反查；内部主机名（jenkins/staging/admin-portal）也可能通过 CT 日志泄露 - **验证**: `curl -H "Host: target.com" https://<候选IP>` 返回目标内容
  • Network accessline 19
    - **假设**: 源站 IP 隐藏但可通过证书/DNS 历史/特征反查；内部主机名（jenkins/staging/admin-portal）也可能通过 CT 日志泄露 - **验证**: `curl -H "Host: target.com" https://<候选IP>` 返回目标内容 - **找 IP 管线**: SecurityTrails DNS 历史 A 记录时间线 → FOFA `cert.serial_number="<hex>"` + Shodan `ss
  • Network accessline 20
    - **验证**: `curl -H "Host: target.com" https://<候选IP>` 返回目标内容
  • Network accessline 55
    - **信号**: 目标有历史资产、旧端点、下线接口 - **假设**: Wayback/Common Crawl 归档了历史 URL 与参数，含已下线端点和带参 URL（`redirect=`/`file=`/`url=` → SSRF/开放重定向线索） - **验证**: `gau target.com`（一次拉 Wayback+Common Crawl+OTX+URLScan）/ `echo target.com | waybackurls`；Common Crawl
  • Network accessline 56
    - **假设**: Wayback/Common Crawl 归档了历史 URL 与参数，含已下线端点和带参 URL（`redirect=`/`file=`/`url=` → SSRF/开放重定向线索） - **验证**: `gau target.com`（一次拉 Wayback+Common Crawl+OTX+URLScan）/ `echo target.com | waybackurls`；Common Crawl CDX `https://index.commoncr
  • Network accessline 57
    - **验证**: `gau target.com`（一次拉 Wayback+Common Crawl+OTX+URLScan）/ `echo target.com | waybackurls`；Common Crawl CDX `https://index.commoncrawl.org/` 用 `*.target.com` 通配挖子域，全程不触达目标
  • Network accessline 85
    - **验证**: 触发目标发信 → 看邮件"原始/源代码" → 读 `Received:` 链最底部（最早）一跳 + `X-Originating-IP`/`X-Sender-IP`（HELO/EHLO 握手常带客户端 LAN IP） - **证伪**: 目标用 SendGrid/Mailgun/阿里云邮件推送 → 拿到的是服务商 IP - **证实**: `curl -H "Host: target.com"` 验证候选 IP 返回目标内容
  • Network accessline 86
    - **证伪**: 目标用 SendGrid/Mailgun/阿里云邮件推送 → 拿到的是服务商 IP - **证实**: `curl -H "Host: target.com"` 验证候选 IP 返回目标内容 - **升级**: 拿到源站 IP → 跳过 WAF 直接探测
  • Network accessline 87
    - **证实**: `curl -H "Host: target.com"` 验证候选 IP 返回目标内容
  • Network accessline 127
    - **信号**: 源站接 CDN，前端 IP 隐藏，但源站在 80/8080/8443 等端口可能仍原样返回同一 body - **假设**: 页面里独一无二的字符串（内联 JS 版本、构建 hash、后台标识）在源站未接 CDN 时全网可搜到；比 favicon 更抗改（有些站改了图标但没改 body） - **验证**: 提取页面唯一特征串 → FOFA `body="<唯一串>"`（去掉 host/domain 限制全网搜）→ 候选 IP 用 `curl -H "Ho
  • Network accessline 128
    - **假设**: 页面里独一无二的字符串（内联 JS 版本、构建 hash、后台标识）在源站未接 CDN 时全网可搜到；比 favicon 更抗改（有些站改了图标但没改 body） - **验证**: 提取页面唯一特征串 → FOFA `body="<唯一串>"`（去掉 host/domain 限制全网搜）→ 候选 IP 用 `curl -H "Host: target.com"` 验证 - **证实**: 命中未接 CDN 的真实源站 IP
  • Network accessline 129
    - **验证**: 提取页面唯一特征串 → FOFA `body="<唯一串>"`（去掉 host/domain 限制全网搜）→ 候选 IP 用 `curl -H "Host: target.com"` 验证

skills/pentest-redteam/techniques/reversing.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash
  • Access to secretsline 33
    [Credential-related source omitted]
  • Access to secretsline 62
    [Credential-related source omitted]
  • Access to secretsline 63
    [Credential-related source omitted]
  • Access to secretsline 64
    [Credential-related source omitted]
  • Access to secretsline 65
    [Credential-related source omitted]

skills/pentest-redteam/techniques/web.md

  • Broad permissionsline 1
    Read,Grep,Glob,Bash,WebFetch

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Signal confidence: 57 facts

and 7 more

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativePath .claude/settings.json.claude/skills/<skill-name>/
CodexCompatibleStandard SKILL.md skills/pentest-redteam/SKILL.md.agents/skills/<skill-name>/
CursorCompatibleStandard SKILL.md skills/pentest-redteam/SKILL.md.cursor/skills/<skill-name>/
Gemini CLICompatibleStandard SKILL.md skills/pentest-redteam/SKILL.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md skills/pentest-redteam/SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md skills/pentest-redteam/SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.

You might also use

mukul975/Anthropic-Cybersecurity-Skills

817 structured cybersecurity skills for AI agents · Mapped to 6 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND, NIST AI RMF & MITRE F3 (Fight Fraud) · agentskills.io …

35.7kGitHub stars

BitterSecurity/Decepticon

Autonomous Hacking Agent for Red Team

6.6kGitHub stars

wgpsec/AboutSecurity

Everything for pentest. | 渗透测试知识库，以 AI Agent 可执行的格式沉淀安全方法论。

1.8kGitHub stars

uphiago/recon-skills

Recon & pentest skill pack. CORS, XSS, SQLi, SSRF, RCE, WordPress, MCP, cloud, subdomain takeover, and more. Field-tested. MIT. Full write-up at hiago.sh

1.3kGitHub stars

transilienceai/communitytools

Open-source Claude Code skills, agents, and slash commands for AI-powered penetration testing, bug bounty hunting, and security research

609GitHub stars

blacklanternsecurity/red-run

Offensive security toolkit for Claude Code

302GitHub stars