The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill takes user-supplied input through the <question> parameter. This input is interpolated into the reasoning logic of the 'Council Coordinator'. While there are no explicit boundary markers to separate user content from the internal instructions, the risk is minimal as the skill lacks any capabilities to write files, execute code, or make network requests.
Ingestion points: user input via the /council command and its variants.
Boundary markers: Absent; the user's question is processed directly by the simulated personas.
Capability inventory: None. The skill does not use any external tools, file operations, or network calls.
Sanitization: None detected for the text input.
[NO_CODE]: The skill is composed entirely of markdown instructions and simulation logic. It does not include Python scripts, Node.js packages, or shell commands, which significantly reduces the attack surface.

council