Works with the AI Agents you already use
AI Agent Security Suite — Scan, Audit, Architect, Respond (4 Skills)
The complete agent security lifecycle in one install: injection scanner, red-team auditor, MCP architect, and incident commander — with a router that chains them.
$29
AI Agent Security Suite — Scan, Audit, Architect, Respond (4 Skills)
Example session with this skill installed
Run the pre-launch security gate on my support agent — 4 tools, Postgres memory, one MCP knowledge base. Ship decision needed Friday.
- Read your context and instructions
- Compiled the ai agent security
Suite routing: multi-stage request → pre-launch gate chain (scanner → auditor).
Stage 1 — prompt-injection-scanner — verdict: SHIP WITH FIXES
| Surface | Probes | Failed | Risk |
| direct injection | 12 | 1 | high |
| extraction | 6 | 1 | medium |
Fix for the two failed surfaces applied and re-verified.
Stage 2 — ai-redteam-auditor — verdict: DO NOT SHIP
Carried scanner findings forward. New: memory-poisoning via ticket body (RT-1, CRITICAL), MCP kb_search returns prose (RT-3, MED). OWASP/ASI matrix: 8 categories ≥90% coverage. CI gate config attached — Friday ship decision: hold until RT-1 fix + re-test (est. same-day loop).
Chain note: findings passed verbatim between stages; auditor verdict governs.
Connects securely to your tools. The creator never sees your data.
About this skill
Four specialist skills, one install — the complete agent security lifecycle: scan prompts for injection, audit the whole system adversarially, architect MCP servers to spec, and command incidents when things break. A built-in router picks the right skill for each request and chains them for multi-stage work.
What's inside
| Skill | Does | Bought alone |
|---|---|---|
| Prompt Injection Scanner | Point-in-time adversarial scan of prompts, skills, and agent configs — 7 probe families, severity scorecard, honest coverage statement | Free |
| AI Red-Team Auditor | Full audit program — injection, jailbreak, leakage, memory poisoning, MCP tool poisoning — OWASP LLM + ASI matrix, false-positive measurement, CI gate config | $19 |
| MCP Server Architect | Designs and reviews MCP servers to current spec — OAuth 2.1 + CIMD, transport selection, tool hash-pinning, per-tool scopes, MRTR | $15 |
| Incident Commander | Structured incident response — SEV0–SEV4, rollback-first, war room, status cadence, timelines, blameless postmortems | $9 |
Bought separately: $43. This suite: $29 — plus the router that chains them (pre-launch gate, MCP integration review, post-incident hardening).
How it works
Ask naturally — "scan this prompt", "red-team my agent", "review this MCP server", "we're having an outage" — and the router loads the right specialist. Or ask for a full pass: "run the pre-launch security gate on my agent" chains scanner → auditor → verdict.
Why a suite
Each skill enforces its own output contract — enum verdicts, evidence-required findings, no bare "looks safe" claims. Together they close the loop most teams leave open: the scanner finds the shallow holes, the auditor proves the deep ones, the architect keeps new integrations clean, and the commander picks up the pieces.
Known limitations
Defensive use only — systems you own or are authorized to test. Skills are instruction packages; they run inside your agent, no external services required. Auto-triggering varies by host agent — explicit invocation (/skill-name) always works.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Defensive use only. Instruction packages running inside your agent — no external services needed. Auto-triggering varies by host agent; explicit invocation always works.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 10 days ago
- Software
- Non MCPMethod MCPName MCPpromptfoo
Node.js
Permissions
No special permissions declared or detected