The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests external specifications to produce implementation plans. These plans are designed to be executed by other agents and include executable components like shell commands (e.g., git, pytest) and Python code blocks. A malicious specification could potentially lead to the generation of a plan containing harmful commands. * Ingestion points: External project specifications or requirements documents referenced in SKILL.md. * Boundary markers: Absent; there are no specific markers used to isolate untrusted input within the generated plan. * Capability inventory: The skill writes plans to the local filesystem and generates instructions for code creation and shell command execution. * Sanitization: The instructions do not include steps for sanitizing or validating the content of the input specifications.

writing-plans