a2a-security-review
Review agent-to-agent (A2A) / multi-agent-system trust: peer identity and authentication, message integrity, capability-negotiation trust, and delegation-chain…
Run an authorized network penetration test: host discovery, port/service enumeration, vulnerability identification, exploitation, and lateral movement/pivoting — within an agreed scope and rules of engagement. Use to actively test internal/external network security you're
$ npx -y skills add jassics/awesome-claude-security --skill network-pentest --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/network-pentestContext preview
The summary Claude sees to decide when to auto-load this skill.
Run an authorized network penetration test: host discovery, port/service enumeration, vulnerability identification, exploitation, and lateral movement/pivoting — within an agreed scope and rules of engagement. Use to actively test internal/external network security you're
name: network-pentest description: >- Run an authorized network penetration test: host discovery, port/service enumeration, vulnerability identification, exploitation, and lateral movement/pivoting — within an agreed scope and rules of engagement. Use to actively test internal/external network security you're authorized to assess.
Evidence-backed network findings and demonstrated impact (where authorized), mapped to a methodology and ranked — without straying outside scope or causing disruption.
window, rules of engagement, and allowed techniques. Do not touch out-of-scope systems. Coordinate to avoid outages.
1. **Host discovery** — identify live hosts within scope (respecting rate limits). 2. **Port & service enumeration** — open ports, service/version fingerprinting, default/again-exposed admin services. 3. **Vulnerability identification** — map services to known weaknesses and misconfigurations; validate, don't just trust a scanner. 4. **Exploitation** (only if authorized, to the depth agreed) — confirm exploitable issues with controlled, non-destructive proof. 5. **Post-exploitation / lateral movement** — assess what the foothold reaches: credential exposure, trust relationships, pivoting — within RoE. 6. **Document** — findings with evidence, impact, and remediation; log what you did for deconfliction.
A findings table: host/service · issue · evidence · impact · remediation, plus a network attack-path view (`security-diagramming:attack-tree`) for notable chains. Confirmed issues → `security-reporting:finding`.
Validate before exploiting and keep proofs non-destructive — this is assessment, not disruption. Stay strictly in scope and log activity for deconfliction. The highest-impact network findings are usually exposed admin services, weak/again-used credentials, and flat networks enabling lateral movement (see `network-segmentation-review`).
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…