M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
A

Agent Scan

Community
Not a standard server
Security scanner for AI agents, MCP servers and agent skills to detect prompt injections and vulnerabilities.

This is not a typical connect-and-use MCP server; read the source repository before using it.

How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #147 of 438
Stars
★ 3.1k Very popular
Transport
stdio (local process)
Runtime
Python · Prebuilt binary
Credentials
API key / credential required
License
Apache-2.0
Last commit
58FMRS · C

Agent Scan is a security scanning tool by Snyk focused on detecting vulnerabilities in AI agents and MCP servers. It supports multiple agents and provides a detailed detection coverage matrix. The tool emphasizes the risk of executing commands during MCP config scanning and recommends sandboxing. Due to experimental CLI output, it is not suitable for automated dependencies.

Strongest · Maintenance 16/20 Weakest · Reliability 7/20

Reliability
7/20
Security and permissions
12/20
Maintenance
16/20
Documentation
15/20
Setup experience
8/20
Why each score
Reliability 7/20
Evidence shows a CLI-first tool rather than an MCP server with a standard MCP lifecycle handshake; no explicit MCP server manifest or verifiable MCP init/tool-list tests are present. The README claims auto-discovery and scanning of MCP servers, and key paths depend on executing arbitrary stdio commands and interacting with a remote Snyk API, but static review cannot confirm these paths reproduce cleanly. Because no verifiable execution evidence is provided, and uncertainty/missing evidence lowers the score, reliability is capped at 7.
Security and permissions 12/20
The tool's security model is thoughtful: it requires explicit user consent before starting each stdio MCP server by default, provides a --dangerously-run-mcp-servers flag to override, and warns clearly to run in a sandbox. Credentials are handled via the SNYK_TOKEN environment variable and no hardcoded real tokens appear in the README. Data flow is disclosed (skills, tool names, and descriptions are sent to the Snyk API). As a scanner, it inherently launches configured servers and communicates with an external API; the inherent risk is real, but the confirmation mechanism, warnings, and disclosure are explicit and substantive. No unmitigated red-line risk is apparent, yet residual risk remains: it executes commands from configs, sends data externally, has a dangerous override flag, and static review cannot verify every permission boundary. Score: 12.
Maintenance 16/20
The repository is maintained by Snyk, has an Apache-2.0 license, and shows significant adoption signals (~2869 stars, only 4 open issues). The README states that external contributions are not accepted but welcomes suggestions and bug reports via GitHub issues. A CHANGELOG.md is present, and mentions of release pipelines and SBOMs indicate ongoing releases. Commit frequency and issue response times cannot be verified statically. Given clear ownership, license, and visible release/changelog evidence, but missing verifiable contribution cadence or security-response channel details, the score is 16.
Documentation 15/20
The README is extensive, covering installation, authentication, CLI parameters, examples, security warnings, a support matrix, and binary verification steps, with links to dedicated CLI reference, scanning guide, JSON output schema, and issue-code documentation. Limitations are explicitly disclosed (experimental CLI output, data sent to Snyk). The documentation is oriented toward CLI users rather than developers wiring this into a mainstream MCP client; there are no MCP server configuration examples, tool parameter references, or troubleshooting sections. Most claims cannot be verified statically, and the docs do not address running this as an MCP server. Score: 15.
Setup experience 8/20
A clear quick-start path is provided: install uv, set SNYK_TOKEN, then run uvx snyk-agent-scan@latest. Standalone binaries are also offered with checksum verification instructions. However, this is not an MCP server that a mainstream client can connect to; it is a CLI that scans MCP configurations. As a CLI, setup is straightforward, but there are no client configuration examples, it requires an external account (Snyk API token), and its installation path is non-standard for MCP server users. Static review cannot verify end-to-end installation; per the calibration cap without execution evidence (max 15), and given these gaps, the score is 8.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessReads local filesRuns commands or codeUses the network

Best for

  • Security teams to monitor company-wide agent supply chain risk
  • Developers to quickly assess security of their MCP servers and skills
  • Evaluating third-party MCP configs in sandboxed environments

Not for

  • Users needing stable programmatic output (CLI output is experimental and may change)
  • Production workflows depending on specific CLI output fields
  • Scanning non-stdio MCP servers (only stdio is supported)

Required permissions

  • Requires SNYK_TOKEN environment variable to call Snyk API
  • Executes commands defined in MCP configurations (stdio server startup)
  • Reads local agent configuration files (e.g., ~/.claude/skills)

Risks and side effects

  • Scanning MCP configurations executes their commands, posing risk of malicious code execution
  • Skills, app names, and tool descriptions are shared with Snyk (usage implies consent)
  • CLI output is experimental and subject to change
  • Large-scale API usage may result in account blocking

Setup

Before you start

Runtime：Python · Prebuilt binary

SNYK_TOKEN requiredsecret Snyk API token, required; get it at https://app.snyk.io/account (API Token → KEY → click to show) and set it via export SNYK_TOKEN=...
  1. Sign up at Snyk and get an API token from https://app.snyk.io/account
  2. Set the token as an environment variable: export SNYK_TOKEN=your-api-token-here
  3. Run with uvx: uvx snyk-agent-scan@latest

Or download a standalone binary from GitHub Releases.

Check that it works

Run uvx snyk-agent-scan@latest; if it auto-discovers local MCP configs/skills and produces a security report without complaining about a missing token, the install works.

Troubleshooting

  1. Ensure SNYK_TOKEN is set correctly, otherwise scans will fail
  2. In non-interactive environments, use --dangerously-run-mcp-servers to bypass consent prompts
  3. If servers time out, adjust --server-timeout
  4. Check that MCP JSON config file paths are correct

Things to try

Once connected, you can ask your AI assistant things like:

  • Scan my machine for MCP configs and agent skills
  • Scan the config file ~/.vscode/mcp.
  • Just inspect tool descriptions without security verification
  • Scan all skills in ~/.claude/skills

Use cases

Scan all installed agent components on your machine for security vulnerabilities
Scan specific MCP configuration files (e.g., ~/.vscode/mcp.json)
Scan individual agent skill files (e.g., SKILL.md) or skill directories
Automate scanning in CI/CD pipelines (requires --dangerously-run-mcp-servers)

Supported clients

Claude Desktop
Claude Code
Cursor
VS Code
Windsurf
Gemini CLI
Amazon Q
Amp
Kiro
OpenCode
Antigravity
Codex
OpenClaw

Listed from the project's documentation, not tested by this site.

Overview

Agent Scan is a security scanning tool that discovers and scans agent components on your machine, including MCP servers and skills, for common threats such as prompt injections, sensitive data handling, or malware payloads hidden in natural language. It automatically discovers configurations for various agents including Claude Code/Desktop, Cursor, Gemini CLI, and Windsurf. The tool supports Scan Mode (CLI scan with report) and Background Mode (periodic scans reported to Snyk Evo). Note: scanning MCP configurations executes the commands defined in them, so it is recommended to run in a sandbox and carefully review consent prompts.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 62.9k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 2d3ca361e334 Data synced 2026-10-12 Read the FMRS scoring method