Token导航 LogoToken导航TokenDH.com
研究检索执行命令github未标认证来源可访问许可证需确认审计异常

pasta-vulns面食创伤

Agent Skill

pasta-vulns 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

272

周安装

11

GitHub Stars

9

下载量

85
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pasta-vulns(面食创伤)
来源仓库:https://github.com/florianbuetow/claude-code
仓库路径:skills/pasta-vulns
安装命令:
npx skills add https://github.com/florianbuetow/claude-code --skill pasta-vulns
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/florianbuetow/claude-code --skill pasta-vulns

简介

用于查找、检索和筛选相关信息。pasta-vulns 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

  • 适合根据关键词或任务场景快速定位候选结果。
  • 可结合来源仓库和原始 README 核验具体用法。
  • 安装前建议确认权限范围和维护状态。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。
  • 注意是否会触发联网或文件读写操作。

SKILL.md

PASTA Stage 5: Vulnerability Analysis

Identify specific weaknesses in code and configuration that could be exploited by Stage 4 threats. This is the core code analysis stage of PASTA. Map each finding to CWE identifiers and correlate with the threat catalog.

Supported Flags

Read ../../shared/schemas/flags.md for the full flag specification. Key behaviors:

FlagStage 5 Behavior
--scopeDefault changed. Analyzes source code, configs, and infrastructure files.
--depth quickScanners + grep patterns only, no manual code review.
--depth standardFull code read, local data-flow analysis, CWE mapping.
--depth deepStandard + cross-file taint analysis, entry-to-sink tracing, dependency CVE scan.
--depth expertDeep + exploitability scoring, proof-of-concept path assessment.
--severityFilter output by severity.
--fixGenerate fix suggestions for each vulnerability found.

Framework Context

Read ../../shared/frameworks/pasta.md, Stage 5 section. PASTA is SEQUENTIAL. Stage 5 consumes Stages 1-4 output and feeds Stage 6.

Prerequisites

Required: Stage 4 output -- threat catalog with MITRE ATT&CK mappings and threat-to-component mapping. Also needs: entry points (Stage 2), components and trust boundaries (Stage 3), business-critical assets (Stage 1). If unavailable, warn and assume.

Workflow

Step 1: Determine Scope

Parse --scope flag (default: changed). Filter to code and config file types. Prioritize files in components targeted by Stage 4 threats.

Step 2: Check for Scanners

ScannerDetectCoverage
semgrepwhich semgrepInjection, auth, crypto, SSRF, XSS
banditwhich banditPython: injection, crypto, subprocess
gosecwhich gosecGo: injection, crypto, file handling
brakemanwhich brakemanRails: injection, XSS, mass assignment
npm auditwhich npmNode.js dependency vulnerabilities
trivywhich trivyContainer and dependency vulnerabilities
gitleakswhich gitleaksSecrets and credentials in code

Step 3: Run Scanners

Run available scanners, normalize output to ../../shared/schemas/findings.md.

Step 4: Manual Code Analysis

  1. Trace data flows: User input from entry points through components to sinks.
  2. Check sanitization: Validation, encoding, parameterization between source and sink.
  3. Review auth/authz: Authentication enforcement and authorization consistency.
  4. Check crypto: Secure algorithms, key management, TLS enforcement.
  5. Review configs: Default credentials, debug modes, security headers.
  6. Check secrets: Hardcoded credentials, API keys, tokens in source.

Step 5: Correlate with Threats

Map each vulnerability to Stage 4 threat(s) it enables: which actor exploits it, which ATT&CK technique it supports, which business asset it endangers.

Step 6: Assess Exploitability

Evaluate: attack complexity, prerequisite access, mitigating controls, and chaining potential with other vulnerabilities.

Analysis Checklist

  1. Are parameterized queries used everywhere, or are there dynamic query paths?
  2. Are there deserialization points accepting untrusted input?
  3. Do all endpoints enforce authentication and authorization?
  4. Are secrets hardcoded or in configuration files within the repository?
  5. Are input validation and output encoding applied consistently?
  6. Are cryptographic algorithms and key lengths secure and current?
  7. Are dependencies up to date with no known CVEs?
  8. Are security headers configured (CSP, HSTS, X-Frame-Options)?

Output Format

Stage 5 produces a Vulnerability Inventory with CWE Mappings. ID prefix: PASTA (e.g., PASTA-001).

Each finding includes: id, title, severity, location (file, line, function, snippet), description, impact, fix, and references (CWE, MITRE ATT&CK, OWASP).

## PASTA Stage 5: Vulnerability Analysis

### Vulnerability Inventory
| ID | Vulnerability | CWE | Severity | Component | Enables Threat |
|----|--------------|-----|----------|-----------|---------------|
| PASTA-001 | SQL injection in search | CWE-89 | Critical | C-02 API | T-01 |
| PASTA-002 | Missing auth on export | CWE-862 | High | C-04 Admin | T-03 |

### Vulnerability-Threat Correlation
| Vulnerability | Threats Enabled | Complexity | Existing Controls |
|--------------|----------------|------------|-------------------|
| PASTA-001 | T-01, T-05 | Low | None |

### Scanner Coverage
| Scanner | Status | Findings |
|---------|--------|----------|
| semgrep | Available / Not found | N findings |

Findings follow ../../shared/schemas/findings.md with:

  • references.cwe: CWE identifier, references.mitre_attck: linked technique, references.owasp: OWASP category
  • metadata.tool: "pasta-vulns", metadata.framework: "pasta", metadata.category: "Stage-5"

Next Stage

Stage 6: Attack Simulation (pasta-attack-sim). Pass the Vulnerability Inventory and threat correlations. Stage 6 constructs exploit chains and scores each attack scenario by exploitability and impact.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.61%
按下载量换算31

Claude

30.64%
按下载量换算26

Cursor

19.86%
按下载量换算17

Gemini CLI

10.08%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

未通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/florianbuetow/claude-code --skill pasta-vulns 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills