The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes curl for performing HTTP operations to create tasks, manage projects, and upload binary file data.
[EXTERNAL_DOWNLOADS]: The skill interacts with external API endpoints at api.manus.ai and manus.im to transmit task data and retrieve agent execution results.
[PROMPT_INJECTION]: There is an inherent surface for indirect prompt injection as the skill processes user-supplied task prompts and project instructions (ingestion points: SKILL.md). These inputs are delimited within JSON structures (boundary markers) and sent to external services via network requests (capabilities: curl network access). While no specific sanitization is mentioned, this behavior is a standard requirement for the skill's primary functionality of agent task automation.

manus