The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the user's message via the <editor /> XML tag, including a project path and element metadata. This data is used to determine which files to read and modify. The skill lacks explicit sanitization or boundary markers to prevent the agent from obeying instructions that might be embedded within the project files or the metadata. The combination of data ingestion with file-writing capabilities and shell command execution creates an attack surface for indirect injection.
[COMMAND_EXECUTION]: The skill instructions require the agent to execute npx maxmotion validate in the project directory. This involves running a CLI tool in the shell, which could be exploited if the project environment or the tool itself is compromised.
[EXTERNAL_DOWNLOADS]: The execution of npx maxmotion triggers a download of the maxmotion package from the public NPM registry. This represents a dependency on external code that is executed at runtime during the skill's operation.

maxmotion-edit