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 the cyber threat intelligence lifecycle for a question or dataset — direction, collection, processing, analysis with structured techniques, and dissemination — to produce an assessed, actionable intelligence product. Use to turn raw threat data into decision-useful
$ npx -y skills add jassics/awesome-claude-security --skill cti-analysis --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/cti-analysisContext preview
The summary Claude sees to decide when to auto-load this skill.
Run the cyber threat intelligence lifecycle for a question or dataset — direction, collection, processing, analysis with structured techniques, and dissemination — to produce an assessed, actionable intelligence product. Use to turn raw threat data into decision-useful
name: cti-analysis description: >- Run the cyber threat intelligence lifecycle for a question or dataset — direction, collection, processing, analysis with structured techniques, and dissemination — to produce an assessed, actionable intelligence product. Use to turn raw threat data into decision-useful intelligence for a defined audience.
An intelligence product that answers a specific requirement, states assessments with calibrated confidence, and tells the consumer what to do — not just a pile of data.
1. **Direction** — define the intelligence requirement and the audience (SOC, IR, leadership). What decision will this inform? This sets relevance. 2. **Collection** — gather from the right sources (internal telemetry/`dfir`, OSINT, feeds, ISAC/sharing, vendor reports). Track source reliability. 3. **Processing** — normalize, deduplicate, translate, and structure (IOCs, TTPs); map to ATT&CK. 4. **Analysis** — apply structured analytic techniques (e.g. Analysis of Competing Hypotheses) to reduce bias; use the Diamond Model and Kill Chain to frame adversary activity; assign confidence and source-reliability ratings. 5. **Dissemination** — deliver in the form/altitude the audience needs (strategic / operational / tactical) with clear, actionable recommendations. 6. **Feedback** — capture whether it met the requirement; refine.
1. Pin the requirement and audience first; scope collection to it. 2. Process and map data to ATT&CK; enrich indicators (`ioc-enrichment`). 3. Analyze with explicit hypotheses and calibrated confidence (distinguish fact vs. assessment vs. assumption). 4. Produce the right product for the audience with recommended actions.
An intelligence product: requirement · key judgments (with confidence) · supporting evidence · ATT&CK/Diamond framing · recommendations · sources (with reliability). Use `security-reporting`; tactical output feeds `detection-engineering`.
Intelligence is requirement-driven and decision-oriented — data without a consumer and a recommendation isn't intelligence. State confidence explicitly and separate fact from assessment from assumption. Match altitude to audience: leadership wants strategic implications, the SOC wants tactical TTPs/IOCs.
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…