The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads task content from external sources (Google Tasks) which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Data from Google Tasks, including task names and notes, enters the agent context via the gws tasks list and gws tasks get commands defined in SKILL.md.
Boundary markers: There are no instructions provided to the agent to treat retrieved task content as data only or to ignore embedded instructions within that content.
Capability inventory: The skill provides the agent with several powerful capabilities, including the ability to delete (delete, clear), modify (update, patch), and move tasks.
Sanitization: The instructions do not define any sanitization, filtering, or validation steps for the content retrieved from the Google Tasks API before it is processed.
[COMMAND_EXECUTION]: The skill is designed to execute commands using the gws CLI tool.
Evidence: The metadata section explicitly requires the gws binary, and the instructions provide templates for various gws tasks subcommands. It also references a shared local file (../gws-shared/SKILL.md) for authentication and global configuration.

gws-tasks