Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

AgentScan

agentscan.sh

  • Free plan
  • Web
  • API
7.4 Club score Visit agentscan.sh

What AgentScan is

AgentScan is a cloud service for automated security testing of LLMs, retrieval-augmented generation pipelines, and autonomous agents through an API. It covers attack patterns such as prompt injection, data exfiltration, jailbreaks, and MCP attacks. Scan results include a 0–100 risk score, severity breakdown, remediation advice, and evidence, with JSON or PDF export. The web dashboard can detect agent formats, show scan progress and errors, adjust throttling, and provide PDF downloads. 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. The compliance documentation says its OWASP mapping does not test LLM04 Model Denial of Service or LLM10 Model Theft. The Free plan costs nothing and includes five scans monthly and 185 attack vectors. Starter costs $29 a month and includes 30 scans monthly, 300 attack vectors, PDF reports, and three concurrent scans. Customers may scan only agents they own or have explicit written authorization to test.

Who it suits

AgentScan suits teams that need automated security scans for LLMs, RAG pipelines, or autonomous agents. Its API and web dashboard support teams connecting scans to workflows and reviewing results.

Good to know, and what to watch

  • Provides risk scores and remediation advice.
  • Exports evidence in JSON or PDF.
  • REST API supports completion webhooks.
  • Free plan includes five scans per month.
  • Maps vulnerabilities to OWASP LLM Top 10 risks.
  • Free plan is limited to five monthly scans.
  • OWASP mapping excludes two listed risks.
  • Scanning requires ownership or written authorization.

Enterprise

—

Free

Free

Starter

$29/mo

Pro

$99/mo

Free plan

The verdict

AgentScan provides scan results, reporting, and API automation for testing AI agents and pipelines. Check its plan limits and the two excluded OWASP-mapped risks, and scan only systems you are authorized to test.

What the maker publishes

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

Checked against the maker's pages on 3 Oct 2026.