The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to install a package named mcp-edd-analytics-vantage via npm and pip. It also directs users to clone a repository from https://github.com/dhapat3927/mcp-edd-analytics-vantage.git. These sources are not associated with a known trusted organization or well-known service.
[COMMAND_EXECUTION]: The skill provides various CLI commands (e.g., commercevault serve, commercevault sync, commercevault export-analytics) that execute shell operations on the local machine and interact with external services, including database synchronization.
[CREDENTIALS_UNSAFE]: The skill requires the configuration of sensitive Easy Digital Downloads credentials (EDD_CONSUMER_KEY, EDD_CONSUMER_SECRET) via environment variables or a .env file. While secret management via environment variables is a standard practice, providing these keys to software from an untrusted source poses a risk of credential exposure.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by ingesting untrusted data from an external WordPress EDD installation.
Ingestion points: Functions like vault.products.list, vault.orders.list, and vault.customers.list in SKILL.md fetch external data from the EDD REST API.
Boundary markers: There are no documented boundary markers or instructions to the agent to ignore embedded commands within the commerce data.
Capability inventory: The skill has the capability to perform database writes (commercevault sync), file exports (commercevault export-analytics), and license generation, which could be abused if malicious instructions are processed from the API data.
Sanitization: There is no evidence of sanitization or filtering of the external API content before it is processed by the agent.

commercevault-edd-commerce-orchestrator