The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches documentation and API references from the official MedusaJS documentation site.
Evidence: https://docs.medusajs.com/resources/references/customer in SKILL.md.
[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to perform web searches and fetch external content, which is then processed as part of the agent's context.
Ingestion points: Multiple Web-search and WebFetch steps in SKILL.md targeting docs.medusajs.com and general search results.
Boundary markers: Absent. There are no explicit instructions to the agent to treat fetched documentation or search results as untrusted data or to ignore embedded instructions.
Capability inventory: The agent is granted Bash, Write, and Edit tools, which could be leveraged if malicious instructions were ingested via external data.
Sanitization: Absent. The instructions do not specify any validation or filtering of the content retrieved from the web before it is used for code generation.

medusa-customers