← ClaudeAtlas

apex-pipelinelisted

Unified application security testing workflow — combines recon, SAST, DAST, manual hunting, validation, and reporting into a single orchestrated pipeline. Runs as an "app" with automated phase transitions. Supports web apps (React, Next.js, Django, Flask, Laravel, Spring, Rails, Express), mobile APIs, GraphQL, REST, gRPC, and microservices. Phases — Phase 0 (target intake + scope lock), Phase 1 (passive recon + tech fingerprint), Phase 2 (SAST deep scan via semgrep/grep/trufflehog), Phase 3 (DAST active probing — nuclei/ffuf/dalfox), Phase 4 (manual hunt — IDOR/SSRF/XSS/SQLi/auth-bypass/race/business-logic/LLM), Phase 5 (chain building + impact escalation), Phase 6 (7-Question Gate validation), Phase 7 (report generation). Use when starting a full security assessment on any application, when asked to "test this app", "audit this codebase", "find bugs in this project", or when you need an end-to-end security workflow that combines static and dynamic analysis with manual expertise.
mlvpatel/sentinel-ai-offensive · ★ 1 · API & Backend · score 60
Install: claude install-skill mlvpatel/sentinel-ai-offensive
# AppSec Workflow — Full Pipeline in One Skill End-to-end application security testing: Intake → Recon → SAST → DAST → Hunt → Chain → Validate → Report. --- ## THE GOLDEN RULE > **Every finding must answer: "What can an attacker steal, break, or take over RIGHT NOW — with proof?"** > > If you can't show a working exploit → it's not a finding. Move on. --- ## PHASE 0: TARGET INTAKE & SCOPE LOCK ### Step 1: Define the Target ``` TARGET_NAME: _______________ TARGET_TYPE: [ ] Web App [ ] API [ ] Mobile Backend [ ] Microservices [ ] Monolith SOURCE_CODE: [ ] Available [ ] Partial (JS bundles) [ ] None (black-box) AUTH_MODEL: [ ] Session Cookie [ ] JWT [ ] OAuth/OIDC [ ] API Key [ ] SAML [ ] None ``` ### Step 2: Lock Scope ``` IN SCOPE: - Domain(s): _______________ - API(s): _______________ - Source: _______________ OUT OF SCOPE: - _______________ RULES: [ ] No DoS / load testing [ ] No social engineering [ ] Rate limit: ___ req/sec [ ] Test accounts only (no real user data) ``` ### Step 3: Crown Jewel Identification Before touching any tool, determine the highest-value targets: | App Type | Crown Jewel | Worst Case | |----------|-------------|------------| | E-commerce | Payment/billing | Drain funds, steal PII | | SaaS | Multi-tenancy | Cross-tenant data access | | Healthcare | Patient data | HIPAA violation, PII leak | | Auth provider | SSO/tokens | Full SSO chain compromise | | FinTech | Transactions | Unauthorized transfers | |