The Agent Skills Directory

[SAFE]: The skill implements a logic-based pipeline for generating documentation such as PRDs and execution plans. It does not perform any sensitive system operations, execute shell commands, or interact with unauthorized external services.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection through its web research functionality.
Ingestion points: Phase 3 (SKILL.md) utilizes WebSearch and WebFetch to ingest untrusted data from the public web for market and competitor analysis.
Boundary markers: No specific boundary markers or "ignore instructions" delimiters are defined for the fetched content.
Capability inventory: The skill's capabilities are limited to writing structured markdown files (questions.md, research.md, PRD.md, plan.md) to a dedicated output directory.
Sanitization: The skill does not specify any sanitization or validation routines for data retrieved via web search before it is used to generate documentation.

idea-os