agent-scan

Command-line security scanner that inventories and scans locally installed agent components (MCP server configurations, MCP tool descriptions, and optionally agent skills) for common agent-supply-chain threats such as prompt injection, tool poisoning/shadowing, toxic flows, malware payloads, untrusted content, and credential/secret handling issues.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ Security security agent-scan mcp modelcontextprotocol supply-chain prompt-injection tooling cli python
⚙ Agent Friendliness
46
/ 100
Can an agent use this?
🔒 Security
46
/ 100
Is it safe for agents?
⚡ Reliability
32
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
40
Documentation
70
Error Messages
0
Auth Simplicity
80
Rate Limits
10

🔒 Security

TLS Enforcement
20
Auth Strength
65
Scope Granularity
20
Dep. Hygiene
55
Secret Handling
70

Security purpose is clear and it states it does not store/log MCP tool call contents/results; it also offers an opt-out flag. However, provided text does not specify transport security details for API calls, scope granularity for tokens, or explicit guidance on safe handling of secrets in logs beyond general statements. Dependencies include mcp[cli] and pyjwt, implying network/auth functionality that should be verified for TLS enforcement and secure defaults.

⚡ Reliability

Uptime/SLA
0
Version Stability
60
Breaking Changes
40
Error Recovery
30
AF Security Reliability

Best When

You need automated, repeatable local checks of agent components (especially MCP servers) to reduce common prompt-injection and tool-related supply-chain risks.

Avoid When

You cannot or do not want any network verification calls (it mentions invoking an Agent Scan API for verification).

Use Cases

  • Inventorying installed agent components (MCP servers/skills) on developer endpoints
  • Detecting prompt injection/tool poisoning/toxic-flow patterns in MCP tool definitions and agent skills
  • Integrating local agent supply-chain security checks into security workflows and audits
  • Monitoring agent supply chain centrally via background mode (MDM/CrowdStrike) and reporting to Snyk Evo

Not For

  • Acting as a full vulnerability scanner for arbitrary codebases unrelated to agent components
  • Guaranteeing that scanned components are safe (it performs pattern/verification checks rather than formal verification)
  • Serving as an interactive agent orchestration runtime (it is a scanner/inspector CLI)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
No
SDK
No
Webhooks
No

Authentication

Methods: Snyk API token via SNYK_TOKEN environment variable (for verification/scan API)
OAuth: No Scopes: No

README indicates signing up for Snyk and using an API token. No OAuth flow or explicit scopes are described in the provided text.

Pricing

Free tier: No
Requires CC: No

Pricing is not specified in the provided content; background mode mentions contacting Snyk for setup and that large-scale API use is considered abuse.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Tooling scans local files and may connect to MCP servers to retrieve tool descriptions; scanning untrusted MCP servers could be risky even if output is suppressed.
  • Background mode sends results to a central Snyk Evo instance; ensure opt-out/storage-file settings match internal policy.
  • Scanning behavior may depend on local install paths and well-known client definitions; components not in those paths may not be discovered.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for agent-scan.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered