a2a-security-review
Review agent-to-agent (A2A) / multi-agent-system trust: peer identity and authentication, message integrity, capability-negotiation trust, and delegation-chain…
Review a host/OS (Linux or Windows) or its baseline image against CIS Benchmark hardening — accounts, services, network, logging, file permissions, and patching. Use to assess server/VM/golden-image hardening you're authorized to review.
$ npx -y skills add jassics/awesome-claude-security --skill host-hardening-review --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/host-hardening-reviewContext preview
The summary Claude sees to decide when to auto-load this skill.
Review a host/OS (Linux or Windows) or its baseline image against CIS Benchmark hardening — accounts, services, network, logging, file permissions, and patching. Use to assess server/VM/golden-image hardening you're authorized to review.
name: host-hardening-review description: >- Review a host/OS (Linux or Windows) or its baseline image against CIS Benchmark hardening — accounts, services, network, logging, file permissions, and patching. Use to assess server/VM/golden-image hardening you're authorized to review.
A hardening assessment of the host or base image against a recognized benchmark (CIS), with findings, severity, and remediation — suitable for building or auditing a golden image.
1. **Accounts & auth** — no unused/default accounts, strong password/lockout policy, no empty passwords, sudo/admin scope, SSH config (no root login, key-based). 2. **Services & packages** — minimize installed packages and running services; disable unneeded daemons; no legacy/insecure services (telnet, ftp, rsh). 3. **Network** — host firewall on with default-deny; disable uncommon protocols; kernel network params (Linux sysctl) hardened. 4. **Filesystem & permissions** — secure mount options, world-writable files, SUID/ SGID review, sensitive file permissions, separate partitions where applicable. 5. **Logging & audit** — auditd/Event Log configured, time sync, log retention and forwarding. 6. **Patching & integrity** — current patch level, automatic updates policy, file integrity monitoring, secure boot where relevant.
1. Identify OS/version and scope; gather config (a CIS-CAT / OpenSCAP / Lynis scan helps) or review the build/image definition. 2. Compare against the matching CIS Benchmark level (L1 baseline / L2 stricter). 3. Record per control: control (CIS ref) · finding · severity · remediation. 4. For images built via IaC, fold fixes back into the build (`iac-security-review`).
A hardening findings table mapped to CIS controls + ranked gaps. Confirmed issues → `security-reporting:finding`.
Harden the **base image** once and deploy it everywhere rather than patching live hosts ad hoc — pair with `iac-security-review` so the hardened build is codified. Pick the CIS level (L1 vs L2) to match the host's role and risk.
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…