The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data in the form of plans, designs, or proposals provided by the user.
Ingestion points: External data enters the agent context through pasted text, file contents, or specified file paths as described in SKILL.md.
Boundary markers: There are no instructions or delimiters defined to separate user data from the skill's own instructions or to ignore embedded commands.
Capability inventory: The skill includes an instruction to save a final summary to a configured documentation location, which involves file-system write operations.
Sanitization: No validation, escaping, or filtering of the processed content is specified before the agent acts upon it.
[NO_CODE]: The skill does not include any scripts, executable files, or package dependencies, relying solely on natural language instructions.

grill-me