The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions and organizational patterns for a multi-agent debate workflow. No security issues were identified within the instructions.
[PROMPT_INJECTION]: The skill defines role-play prompts for subagents (Architect, Skeptic, Pragmatist, Critic), but these are strictly scoped to the decision-making task and do not attempt to bypass safety filters or override global agent constraints.
[DATA_EXFILTRATION]: The skill explicitly forbids writing to non-standard or 'shadow' paths like hidden notes directories. It recommends using standard, authorized integration tools (like GitHub or Linear) only when a decision materially changes the project state.
[NO_CODE]: There are no scripts, binaries, or automated command-line execution patterns included in this skill.

council