The Agent Skills Directory

EXTERNAL_DOWNLOADS (MEDIUM): The installation instructions in rules/getting-started.md require the user to run npx @x402scan/mcp install. This dependency is from a non-trusted npm scope, making it an unverifiable third-party package.
REMOTE_CODE_EXECUTION (MEDIUM): The use of npx to install the @x402scan/mcp package involves downloading and executing code from an external registry at runtime. Without a verified source or organization, this command could lead to arbitrary code execution on the user's system.
DATA_EXFILTRATION (LOW): The skill is designed to send user identifiers (such as emails, names, and LinkedIn URLs) to enrichx402.com for processing. While this is the intended purpose of the skill, the domain is not on the whitelist of trusted sources, and users should be aware of the data shared with this third party.
PROMPT_INJECTION (LOW): The skill processes untrusted data fetched from external APIs, creating a surface for indirect prompt injection (Category 8c).
Ingestion points: API response data retrieved from enrichx402.com via the x402.fetch tool.
Boundary markers: None identified; the instructions direct the agent to "Parse and present results" without specifying safety delimiters or instructions to ignore embedded commands.
Capability inventory: The skill utilizes network access tools (x402.fetch) and API discovery tools (x402.discover_api_endpoints).
Sanitization: No sanitization, validation, or escaping of the external API content is mentioned in the provided files.
COMMAND_EXECUTION (MEDIUM): The skill requires the execution of shell commands (npx) to set up the environment, which is a potential vector for system compromise if the package is malicious.

data-enrichment