The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the '@pnp/cli-microsoft365' package globally using npm. This package is the primary tool for the skill and is provided by the designated vendor.\n- [COMMAND_EXECUTION]: The skill relies on executing various shell commands via the 'm365' CLI to manage Microsoft 365 resources, including SharePoint sites, Teams channels, and Entra ID users.\n- [CREDENTIALS_UNSAFE]: The documentation includes examples of authentication using command-line arguments for passwords and secrets (e.g., '--password', '--secret'). While a best practice note advises against hardcoding secrets, the presence of these flags in examples could lead to users or the agent passing sensitive information in a way that is recorded in system logs or shell history.\n- [INDIRECT_PROMPT_INJECTION]: The skill provides the ability to read content from a Microsoft 365 tenant, such as file contents, list items, or chat messages. This content could contain malicious instructions designed to influence the agent's behavior. The skill lacks instructions for establishing security boundaries or sanitizing data retrieved from these external sources.\n
Ingestion points: Operations like 'm365 spo listitem list', 'm365 teams message list', and 'm365 request' retrieve data from the tenant.\n
Boundary markers: None provided to distinguish between system instructions and tenant-provided data.\n
Capability inventory: Significant write and administrative capabilities, including creating sites, sending messages, and modifying application permissions.\n
Sanitization: No procedures mentioned for validating or escaping content retrieved from the Microsoft 365 environment.

cli-microsoft365