The Agent Skills Directory

[SAFE]: The skill implements a logical framework for strategic decision-making and risk assessment. No malicious behaviors were detected. It defines local file paths (e.g., ~/.claude/memory-palace/strategeion/checkpoints/) for logging audit trails, which is consistent with its stated session persistence and integrity verification functionality.
[PROMPT_INJECTION]: The skill ingests external context such as decision descriptions and conflict summaries to calculate risk scores. This represents an indirect prompt injection surface that is functionally integral to the skill's purpose and does not include instructions to override agent safety protocols or bypass constraints. * Ingestion points: decision_needed, files_affected, issues_involved, blocking_items, conflict_description (SKILL.md) * Boundary markers: None explicitly defined for interpolated context. * Capability inventory: The skill evaluates risk to recommend or trigger the invocation of the attune:war-room skill. * Sanitization: No explicit sanitization or filtering of input context is described.

war-room-checkpoint