Security Expert Skill

Quick Start

bash

# 1. Navigation rapide vers un agent
security-expert/agents/appsec/sast              # Analyse statique du code
security-expert/agents/appsec/dast              # Tests dynamiques
security-expert/agents/secure-coding/validation # Validation des entrees
security-expert/agents/threat-modeling/stride   # Methodologie STRIDE
security-expert/agents/penetration/owasp-top10  # Tests OWASP Top 10
security-expert/agents/compliance/rgpd          # Implementation RGPD

# 2. Executer les tests de validation
cd .web-agency/skills/security-expert && npm test

# 3. Questions frequentes
"Comment implementer SAST dans ma CI ?"     -> appsec/sast
"Valider les entrees utilisateur ?"         -> secure-coding/validation
"Faire un threat model de mon app ?"        -> threat-modeling/stride
"Tester les vulnerabilites OWASP ?"         -> penetration/owasp-top10
"Implementer le RGPD techniquement ?"       -> compliance/rgpd

Position dans l'Architecture

Ce skill est au NIVEAU 3 : IMPLEMENTATION. Il implemente les politiques de securite definies par direction-technique/securite.

code

+---------------------------------------------------------------------------+
|  NIVEAU 1 : STRATEGIE (direction-technique/securite)                       |
|  -> POURQUOI : Politique securite, exigences conformite, budget            |
+----------------------------------------------------------------------------+
|  NIVEAU 2 : PROCESSUS                                                      |
|  +----------------------------+  +----------------------------+            |
|  |     testing-process        |  |       web-dev-process      |            |
|  |  security/ (methodologie)  |  |  testing/security/         |            |
|  |  QUOI : Quels tests faire  |  |  QUOI : Quand tester       |            |
|  +----------------------------+  +----------------------------+            |
+----------------------------------------------------------------------------+
|  NIVEAU 3 : IMPLEMENTATION                                                 |
|  +------------------------------------------------------------------+     |
|  |                    security-expert <- CE SKILL                    |     |
|  |  COMMENT : SAST, DAST, secure coding, pentest, compliance impl    |     |
|  +------------------------------------------------------------------+     |
+----------------------------------------------------------------------------+

Philosophie

Securiser par design, valider en continu.

Ce skill :

•Implemente les outils de securite (SAST, DAST, IAST)
•Guide le secure coding avec des patterns concrets
•Realise les threat models et analyses de risques
•Execute les tests de penetration
•Implemente la conformite (RGPD, SOC2, ISO27001)

Il ne fait PAS :

•Les decisions strategiques de securite -> direction-technique/securite
•La definition des process de tests securite -> testing-process/security
•La gestion des incidents -> direction-technique/support/gestion-incidents
•La configuration infrastructure securisee -> devops/infrastructure

Domaines et Agents (24 agents)

1. appsec/ - Application Security (5 agents)

Outils et implementation de la securite applicative.

Agent	Responsabilite	Technologies
`orchestrator`	Coordination AppSec	-
`sast`	Analyse statique du code	SonarQube, Semgrep, CodeQL
`dast`	Tests dynamiques	OWASP ZAP, Burp Suite, Nuclei
`iast`	Instrumentation runtime	Contrast, Hdiv
`sca`	Analyse des dependances	Snyk, npm audit, Dependabot

2. secure-coding/ - Developpement Securise (5 agents)

Patterns et pratiques de code securise.

Agent	Responsabilite	Technologies
`orchestrator`	Coordination secure coding	-
`validation`	Validation des entrees	Zod, Joi, express-validator
`authentication`	Auth securisee	JWT, OAuth2, OIDC, MFA
`authorization`	Controle d'acces	RBAC, ABAC, policies
`cryptography`	Chiffrement et hachage	bcrypt, argon2, AES, RSA

3. threat-modeling/ - Modelisation des Menaces (4 agents)

Identification et analyse des risques.

Agent	Responsabilite	Technologies
`orchestrator`	Coordination threat modeling	-
`stride`	Methodologie STRIDE	Diagrammes, matrices
`attack-trees`	Arbres d'attaque	Threat modeling tools
`risk-assessment`	Evaluation des risques	CVSS, matrices de risques

4. penetration/ - Tests de Penetration (5 agents)

Tests offensifs et identification de vulnerabilites.

Agent	Responsabilite	Technologies
`orchestrator`	Coordination pentest	-
`owasp-top10`	Tests OWASP Top 10	Injection, XSS, CSRF, etc.
`api-security`	Securite des APIs	OWASP API Top 10
`web-vulnerabilities`	Vulnerabilites web	SQLi, XSS, SSRF, IDOR
`reporting`	Rapports de pentest	Templates, CVSS, remediation

5. compliance/ - Conformite (5 agents)

Implementation des standards de conformite.

Agent	Responsabilite	Technologies
`orchestrator`	Coordination conformite	-
`rgpd`	Implementation RGPD	Consentement, DPO, droits
`soc2`	Conformite SOC2	Controles, evidence, audit
`iso27001`	Implementation ISO 27001	SMSI, controles, certification
`pci-dss`	Conformite PCI DSS	Cartes de paiement, tokenization

Total : 24 agents specialises

Regles de Routage

Par Type de Question

Question	Domaine
SAST, DAST, analyse de code, scan	`appsec/`
Validation, auth, chiffrement, code securise	`secure-coding/`
Threat model, STRIDE, risques, attaques	`threat-modeling/`
Pentest, OWASP, vulnerabilites, injection	`penetration/`
RGPD, SOC2, ISO27001, conformite	`compliance/`

Par Mots-Cles

Mots-cles	Domaine/Agent
SonarQube, Semgrep, CodeQL, analyse statique	`appsec/sast`
ZAP, Burp, scan dynamique, fuzzing	`appsec/dast`
Snyk, npm audit, dependances, CVE	`appsec/sca`
input validation, sanitize, escape, Zod	`secure-coding/validation`
JWT, OAuth, session, MFA, auth	`secure-coding/authentication`
RBAC, permissions, policies, access control	`secure-coding/authorization`
bcrypt, hash, encrypt, AES, RSA, salt	`secure-coding/cryptography`
STRIDE, threat model, DFD, trust boundary	`threat-modeling/stride`
attack tree, kill chain, threat actor	`threat-modeling/attack-trees`
CVSS, risk matrix, impact, likelihood	`threat-modeling/risk-assessment`
OWASP Top 10, injection, XSS, CSRF	`penetration/owasp-top10`
API security, broken auth, mass assignment	`penetration/api-security`
SQLi, SSRF, IDOR, path traversal	`penetration/web-vulnerabilities`
RGPD, GDPR, consentement, DPO, droits	`compliance/rgpd`
SOC2, audit trail, evidence, controls	`compliance/soc2`

Arbre de Decision

code

Requete Security
|
+-- Concerne les outils de scan/analyse ?
|   +-- Analyse statique du code -> appsec/sast
|   +-- Tests dynamiques -> appsec/dast
|   +-- Runtime instrumentation -> appsec/iast
|   +-- Dependances vulnerables -> appsec/sca
|
+-- Concerne le code securise ?
|   +-- Validation entrees -> secure-coding/validation
|   +-- Authentification -> secure-coding/authentication
|   +-- Autorisation -> secure-coding/authorization
|   +-- Chiffrement -> secure-coding/cryptography
|
+-- Concerne l'analyse des risques ?
|   +-- Threat modeling STRIDE -> threat-modeling/stride
|   +-- Arbres d'attaque -> threat-modeling/attack-trees
|   +-- Evaluation risques -> threat-modeling/risk-assessment
|
+-- Concerne les tests de penetration ?
|   +-- OWASP Top 10 -> penetration/owasp-top10
|   +-- Securite API -> penetration/api-security
|   +-- Vulns web specifiques -> penetration/web-vulnerabilities
|   +-- Rapports pentest -> penetration/reporting
|
+-- Concerne la conformite ?
|   +-- RGPD/GDPR -> compliance/rgpd
|   +-- SOC2 -> compliance/soc2
|   +-- ISO 27001 -> compliance/iso27001
|   +-- PCI DSS -> compliance/pci-dss
|
+-- Decision strategique securite ?
|   +-- -> direction-technique/securite
|
+-- Methodologie de test securite ?
    +-- -> testing-process/security

Interaction avec les Autres Skills

Flux Entrants

code

direction-technique/securite --> security-expert (politique -> implementation)
testing-process/security --> security-expert (methodologie -> outils)
web-dev-process/testing --> security-expert (phase testing -> execution)

Flux Sortants

code

security-expert --> backend-developer (patterns securises backend)
security-expert --> frontend-developer (patterns securises frontend)
security-expert --> devops (integration CI/CD securite)

Points d'Escalade

Vers direction-technique

Situation	Raison
Vulnerabilite critique trouvee	Decision remediation urgente
Non-conformite majeure	Impact legal/business
Choix d'outils securite	Decision strategique
Budget securite	Validation financiere

Vers l'humain

Situation	Raison
Pentest sur prod	Autorisation explicite requise
Fuite de donnees suspectee	Responsabilite legale
Choix architectural securite	Impact business
Tests destructifs	Risque operationnel

Security Principles

Defense in Depth

code

+------------------------------------------+
|              WAF / CDN                   |
|  +------------------------------------+  |
|  |         Load Balancer              |  |
|  |  +------------------------------+  |  |
|  |  |    Application (validated)   |  |  |
|  |  |  +-----------------------+   |  |  |
|  |  |  |   Business Logic     |   |  |  |
|  |  |  |  +----------------+  |   |  |  |
|  |  |  |  | Data (encrypted)|  |   |  |  |
|  |  |  |  +----------------+  |   |  |  |
|  |  |  +-----------------------+   |  |  |
|  |  +------------------------------+  |  |
|  +------------------------------------+  |
+------------------------------------------+

OWASP Top 10 Coverage

#	Vulnerability	Agent(s)
A01	Broken Access Control	`secure-coding/authorization`, `penetration/owasp-top10`
A02	Cryptographic Failures	`secure-coding/cryptography`
A03	Injection	`secure-coding/validation`, `penetration/web-vulnerabilities`
A04	Insecure Design	`threat-modeling/stride`
A05	Security Misconfiguration	`appsec/sast`, `devops/containers/security`
A06	Vulnerable Components	`appsec/sca`
A07	Auth Failures	`secure-coding/authentication`
A08	Software & Data Integrity	`appsec/sast`, `devops/cicd`
A09	Logging & Monitoring	`devops/monitoring`
A10	SSRF	`penetration/web-vulnerabilities`

Skills Associes

Skill	Niveau	Relation
`direction-technique`	STRATEGIE	Recoit les politiques securite
`testing-process`	PROCESSUS	Suit la methodologie tests
`devops`	IMPLEMENTATION	Integration CI/CD
`backend-developer`	IMPLEMENTATION	Patterns securises backend
`frontend-developer`	IMPLEMENTATION	Patterns securises frontend

Changelog

v1.0.0

•Creation initiale avec 5 domaines et 24 agents
•Position : NIVEAU 3 IMPLEMENTATION
•Couverture : appsec, secure-coding, threat-modeling, penetration, compliance