a2a-security-review
Review agent-to-agent (A2A) / multi-agent-system trust: peer identity and authentication, message integrity, capability-negotiation trust, and delegation-chain…
Assess a web application against the OWASP Top 10 (2021), producing a per-category finding set with severity and remediation. Use when reviewing or pentesting a web app for the most common, highest-impact web risks. Authorized testing only.
$ npx -y skills add jassics/awesome-claude-security --skill owasp-web-top10 --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/owasp-web-top10Context preview
The summary Claude sees to decide when to auto-load this skill.
Assess a web application against the OWASP Top 10 (2021), producing a per-category finding set with severity and remediation. Use when reviewing or pentesting a web app for the most common, highest-impact web risks. Authorized testing only.
name: owasp-web-top10 description: >- Assess a web application against the OWASP Top 10 (2021), producing a per-category finding set with severity and remediation. Use when reviewing or pentesting a web app for the most common, highest-impact web risks. Authorized testing only.
A structured assessment across the ten OWASP Web categories, each with applicability, evidence, severity, and remediation.
1. **Map the app** — entry points, roles, auth model, sensitive functions, tech stack. A `threat-modeling:stride` pass helps prioritize. 2. **Walk each category** (see `reference.md` for checks + fixes): A01 Broken Access Control · A02 Cryptographic Failures · A03 Injection · A04 Insecure Design · A05 Security Misconfiguration · A06 Vulnerable & Outdated Components · A07 Identification & Authentication Failures · A08 Software & Data Integrity Failures · A09 Security Logging & Monitoring Failures · A10 Server-Side Request Forgery (SSRF). 3. **Substantiate** rather than assert: use `access-control-test` and `injection-test` for those categories; for components, cross-ref `sast-sca`. 4. **Score** each finding (`security-reporting:cvss`) and rank.
A per-category table (category · applicable? · finding · severity · remediation) plus a ranked top-risks list. Confirmed issues → `security-reporting:finding`.
A01 Broken Access Control is the most prevalent category — give it explicit, tested attention. Read `reference.md` for the per-category checklist. Stay within authorized scope.
A Claude Code plugin marketplace for the full cybersecurity & GenAI-security lifecycle — from recon and threat modeling to detection engineering, GRC, and CISO-level strategy. A pentester knows which OWASP test bends a broken-access-control endpoint.
Repo: jassics/awesome-claude-security
Review agent-to-agent (A2A) / multi-agent-system trust: peer identity and authentication, message integrity, capability-negotiation trust, and delegation-chain…
Test the agent execution harness/runtime itself — LangChain/LangGraph, AutoGen, CrewAI, custom ReAct-style loops, or computer-use/browser-use agents — for…
Assess an autonomous / tool-using AI agent for security end-to-end: tool privileges, autonomy and approval boundaries, excessive agency, memory/state…
Test what an AI agent will actually do without human confirmation, including under injected-goal / prompt-injection scenarios, to validate its autonomy and…
Review the security of MCP (Model Context Protocol) servers/clients an agent uses: server trust tier, tool/resource description and result poisoning,…
Inventory the tools/functions an AI agent can call and audit their privileges, side effects, and approval requirements to find excessive-agency and…