The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local .env files which could potentially contain malicious instructions intended to influence the agent's behavior.
Ingestion points: The skill reads files at {workspace_root}/.env or other user-provided paths using a file reading tool in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when parsing the content of .env files before passing the data to other tools.
Capability inventory: The skill utilizes MCP tools to create environments, append variables, and create local file mounts (named pipes) on the filesystem, and performs authentication via the 1Password MCP server.
Sanitization: No explicit sanitization or validation of the .env content is performed before passing data to the 1Password MCP tools.

1password-environments