MCP Registry

@mseep/mcp-inscription

73414510-9453-5600-9e6b-884ba310a53d | Store v0.1.0
64/ 100
MEDIUM risk
AI review verdict
Review before use

The AI review rates the findings as likely false positive, but the risk score (64/100) still counts them.

Analysis record

Analysed
7 months ago 23 April 2026
Version
v0.1.0
Artifact
SHA256 5e60a227…88aecf
Detail
Full list

Evidence ledger

Ranked by severity · findings with a source location link to the code viewer

9 findings

Finding Categories

3
Network

MCP Server Analysis

MCP servers expose tools and resources to AI assistants. Unlike browser extensions, they run as standalone processes with direct system access. Tool definitions are analyzed for prompt injection, data exfiltration, and tool poisoning patterns.

AI Security Report

AI Security Review

Evidence context: threat category none; evidence quality strong.

This MCP server package shows no evidence of malicious behavior. The findings summary explicitly reports zero tool-poisoning findings ("tool-poisoning":"0"), zero malware-signature findings, and zero secret/credential-access findings. The three high-severity filesystem-access findings in build/servers/base.js (lines 351, 381, 395) are labeled "MCP tool poisoning risk" by the scanner, but these are legitimate MCP tool definitions for an inscription server that writes files to the blockchain, not hidden AI manipulation directives. Real tool poisoning contains instructions like "do not tell the user" or "silently execute"—none of which appear in the evidence.

The network findings show two fetch calls in build/ordinals_client.js (lines 277, 331) and a no-auth transport in build/servers/sse.js. The fetch calls align with the package's stated purpose as an "MCP-Inscription Server" for blockchain ordinals, not exfiltration to unknown domains. There are zero IoC findings, meaning no suspicious external domains were detected. The no-auth transport is a security weakness but not malicious exfiltration architecture.

Credential scope is absent: the findings summary shows zero secret findings, and no credential-access findings target sensitive paths like .ssh/, .aws/credentials, or .kube/config. The package reads no environment variables for secrets.

The strongest counterargument is the three high-severity filesystem-access findings labeled "MCP tool poisoning risk." However, these are false positives from the scanner flagging legitimate MCP tool definitions. The findings summary confirms zero actual tool-poisoning detections. An inscription server legitimately needs filesystem access to prepare data for blockchain inscription. The combination of zero tool-poisoning, zero credential access, zero IoC, and zero malware signatures with network calls to ordinals-related endpoints confirms this is a legitimate MCP server with scanner noise, not malicious behavior.

Key Reasons

  • Zero tool-poisoning findings in findings summary
  • Zero credential-access or secret findings
  • Zero IoC findings (no suspicious domains)
  • Network calls align with stated inscription server purpose
  • Filesystem access is legitimate for blockchain inscription functionality

False Positive Considerations

  • MCP tool definition patterns flagged as tool-poisoning risk
  • Filesystem access findings from legitimate inscription server functionality
  • Network fetch calls to ordinals/blockchain endpoints misclassified as suspicious

Reviewed 2026-04-27 · version 0.1.0; recommended action: suppress false positive; model confidence 85%.

Source Code Viewer

Source Code Not Available

Source code is not available for this version of the extension.

About This Extension

MCP-Inscription Server

Frequently Asked Questions