mcp-skill-sec MCP
A deterministic, no-LLM MCP server that audits agent skills, system prompts, and downloaded files against 8 malicious-skill supply-chain patterns including prompt injection, data exfiltration, hardcoded secrets, dangerous commands, obfuscation, untrusted fetches, credential access, and privilege escalation. It provides line-numbered evidence and severity ratings for findings.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Audit agent skills and system prompts for malicious patterns
- Scan downloaded file collections for security risks
- Detect hardcoded secrets and credentials in code
- Identify dangerous commands and privilege escalation attempts
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
mcp-skill-sec 1.0.0
{
"version": "1.0.0",
"transport": "stdio",
"identifier": "mcp-skill-sec",
"registryType": "pypi"
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"mcp-skill-sec-mcp": {
"command": "uvx",
"args": [
"mcp-skill-sec==1.0.0"
]
}
}
}Cline
{
"mcpServers": {
"mcp-skill-sec-mcp": {
"command": "uvx",
"args": [
"mcp-skill-sec==1.0.0"
]
}
}
}VS Code
{
"servers": {
"mcp-skill-sec-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"mcp-skill-sec==1.0.0"
]
}
}
}Claude Desktop
{
"mcpServers": {
"mcp-skill-sec-mcp": {
"command": "uvx",
"args": [
"mcp-skill-sec==1.0.0"
]
}
}
}Let your agent find its tools
Find servers like mcp-skill-sec MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.