The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses instructional language intended to override default conversational biases (e.g., 'override default patterns that favor agreement'). These are behavioral alignment instructions aimed at improving reasoning quality and do not attempt to bypass safety or security filters.
[INDIRECT_PROMPT_INJECTION]: The skill's analysis protocols process user-provided data such as ethical claims and conflict descriptions. Ingestion points: User queries and claims analyzed in modules/conflict-analysis.md. Boundary markers: Analysis is conducted using structured checklist templates. Capability inventory: No tools, network operations, or file-write capabilities are requested or used by this specific skill. Sanitization: No specific data sanitization or escaping is mentioned. The lack of associated code or external tools minimizes the risk of indirect injection.
[NO_CODE]: The skill consists entirely of markdown-based instructions and metadata, with no associated scripts, binaries, or automated tool configurations.

rigorous-reasoning