The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references research and framework guidelines from recognized professional organizations, including McKinsey & Company, Harvard Business Review, Boston Consulting Group, and Strategyzer. These references are used solely for strategic grounding and originate from well-known, trusted domains.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow to analyze user-supplied portfolio data and external strategic articles fetched via search tools.
Ingestion points: Processes user-provided descriptions of business initiatives, strategic ambitions, and content retrieved via WebFetch and WebSearch.
Boundary markers: The skill does not define specific delimiters to separate user data from instructions, relying on standard model reasoning.
Capability inventory: The execution environment is restricted to information retrieval tools (Read, Grep, WebSearch, WebFetch). No high-risk tools such as shell access, file-writing, or code execution are permitted.
Sanitization: No explicit content sanitization is described for the ingested data.
[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. The skill's configuration and code are consistent with its stated purpose of providing a strategic reasoning framework.

three-horizons