AgentScan

Free plan APIWeb

What AgentScan is

AgentScan is a cloud service that tests LLMs, retrieval-augmented generation pipelines, and autonomous agents through an API. It checks attack patterns such as prompt injection, data exfiltration, jailbreaks, and MCP attacks. Results can include a 0–100 risk score, severity breakdown, remediation advice, and evidence exported as JSON or PDF. The web dashboard can detect agent formats, display scan progress and errors, adjust throttling, and download reports. Its REST API queues scans asynchronously, returns a scan ID, and supports completion webhooks. AgentScan maps vulnerabilities to OWASP LLM Top 10 risks and describes mappings to EU AI Act articles and NIST AI RMF functions. Its compliance documentation notes that its OWASP mapping does not test Model Denial of Service or Model Theft. The free plan allows five scans per month with 185 attack vectors; Starter costs $29 a month and includes 30 scans monthly. The terms limit scans to agents customers own or have explicit written authorization to test.

Who it suits

AgentScan suits teams assessing LLMs, RAG pipelines, or autonomous agents and automating scans through an API. Users need ownership or explicit written permission to test the agents they scan.

What it does well

Results include scores, severity, and remediation advice.
Exports evidence in JSON or PDF.
REST API supports completion webhooks.
Free plan includes five scans monthly.
Maps risks to OWASP and other frameworks.

What to know first

Free plan covers 185 attack vectors.
OWASP mapping omits two listed risk tests.
Scans require ownership or written authorization.
Starter costs $29 a month.

Limits

The Free plan allows five scans per month and covers 185 attack vectors across five categories, with JSON results. Starter raises the allowance to 30 scans per month and 300 attack vectors, and adds PDF reports, a full test audit trail, and three concurrent scans.

Plans

Plan	Billed and limits	Price
Free	$0/month · 5 scans per month · 185 attack vectors · 5 categories · auto-detect agent format · JSON results	Free
Starter	$29/month · 30 scans per month · 300 attack vectors · 12 categories · PDF reports · full test audit trail · 3 concurrent scans	$29 a month
Pro	$99/month · 100 scans per month · 348 attack vectors · 14 categories · custom secrets monitoring · configurable throttling · PDF reports + webhooks · priority support	$99 a month
Enterprise	Custom · Unlimited scans · all 412 attack vectors · all 19 categories · MCP, State Injection, PII · SLA + priority support	On request

As agentscan.sh lists them, read 3 Oct 2026. The maker's own page has the final word.

Getting started

Visit the AgentScan website.
Choose the Free plan or a paid plan.
Connect an agent using a documented integration or supported agent format.
Submit a scan through the API or use the web dashboard.
Review the results and export them as JSON or PDF where your plan allows.

Verdict

AgentScan provides API-based security scans with report exports and a free monthly allowance. Note the stated OWASP coverage omissions and authorization requirement when planning assessments.

Questions

Is AgentScan free?

Yes. The Free plan is $0 per month and includes five scans per month, 185 attack vectors, five categories, agent format auto-detection, and JSON results.

How much do paid plans cost?

Starter is $29 per month and Pro is $99 per month. Enterprise is listed at a custom price.

Which platforms does AgentScan support?

The listed platforms are API and web. It is a cloud deployment.

What kinds of agents can it scan?

Documentation lists 12 agent types, including generic HTTP, OpenAI-compatible, Dify, Flowise, n8n, OpenClaw, AutoGPT, OpenHands, Anthropic-compatible, LangChain, and Manus.

Can AgentScan test all OWASP LLM risks?

Its compliance documentation says LLM04 Model Denial of Service and LLM10 Model Theft are not tested in its OWASP mapping.

What authorization is required to scan an agent?

The terms say customers may scan only agents they own or have explicit written authorization to test.

Facts from AgentScan's own pages

Agent support: The documentation lists 12 agent types, including generic HTTP, OpenAI-compatible, Dify, Flowise, n8n, OpenClaw, AutoGPT, OpenHands, Anthropic-compatible, LangChain, and Manus. · agentscan.sh, 3 Oct 2026
Authorized use: The terms say customers may scan only agents they own or have explicit written authorization to test. · agentscan.sh, 3 Oct 2026
Automation: The REST API queues scans asynchronously, returns a scan ID, and supports completion webhooks. · agentscan.sh, 3 Oct 2026
Coverage: The product says it tests 412 attack vectors across 19 categories, including prompt injection, data exfiltration, jailbreak, and MCP attacks. · agentscan.sh, 3 Oct 2026
Coverage limits: The compliance documentation says LLM04 Model Denial of Service and LLM10 Model Theft are not tested in its OWASP mapping. · agentscan.sh, 3 Oct 2026
Dashboard: The web dashboard supports agent-format auto-detection, real-time scan progress and error tracking, live throttle adjustment, and PDF downloads. · agentscan.sh, 3 Oct 2026
Data handling: The privacy policy says scan results remain in the account until deleted and account data is permanently removed within 30 days after account deletion. · agentscan.sh, 3 Oct 2026
Framework mapping: AgentScan maps vulnerabilities to OWASP LLM Top 10 risks and describes mappings to EU AI Act articles and NIST AI RMF functions. · agentscan.sh, 3 Oct 2026
Hosting: The privacy policy says user data is stored on Hetzner servers in Germany (EU). · agentscan.sh, 3 Oct 2026
Integrations: Documented connection examples include OpenAI, Groq, LiteLLM, Dify, n8n, Flowise, OpenClaw, AutoGPT, and OpenHands. · agentscan.sh, 3 Oct 2026
Purpose: AgentScan provides automated pentesting for LLMs, RAG pipelines, and autonomous agents through an API. · agentscan.sh, 3 Oct 2026
Reports: Scan results include a 0–100 risk score, severity breakdown, remediation advice, and JSON or PDF export with evidence. · agentscan.sh, 3 Oct 2026
Security: The privacy policy says passwords are hashed with Argon2id, API keys are stored as HMAC-SHA256 hashes, connections use TLS, and outbound requests have SSRF protection. · agentscan.sh, 3 Oct 2026
Support: The Pro plan lists priority support, while Enterprise lists dedicated support, an SLA, and priority support. · agentscan.sh, 3 Oct 2026

Where AgentScan stands on Freedom251

AI Security Testing Tools#7 of 29