The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run shell commands using pnpm (specifically typecheck, lint, and test) to verify the output of assigned tasks. These are standard development tools used for project maintenance.
[PROMPT_INJECTION]: The skill is subject to potential indirect prompt injection (Category 8) because it ingests untrusted user task descriptions and incorporates them into the tasks.md project file. This content is subsequently used to direct the behavior of other agents.
Ingestion points: User-provided messages containing multiple tasks (Phase 1, SKILL.md).
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the generated tasks.md format.
Capability inventory: The workflow involves shell command execution through the pnpm toolset (Phase 4 and 5, SKILL.md).
Sanitization: The instructions do not specify any sanitization or validation of the user input before it is written to the task list.

task-manager