doc-bootstrapping
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs legitimate project initialization tasks by creating a structured documentation hierarchy including directories for explanations, specifications, how-to guides, and tutorials. It populates these with standard markdown templates.
- [INDIRECT_PROMPT_INJECTION]: The skill establishes agent-specific instructions by writing a template to
.agent/rules/rule-one.md. This path is used by certain agent platforms to define persistent behavioral constraints and protocols. While this represents a mechanism for modifying agent instructions, the content provided in the templates is limited to benign architectural principles, coding standards, and AI-collaboration protocols. - Ingestion points: Static templates located in the
resources/directory (e.g.,rule-one-template.md). - Boundary markers: None; the content is explicitly intended to serve as the agent's core rule set.
- Capability inventory: File-writing capabilities used to create documentation files and rule definitions.
- Sanitization: None; the skill performs direct file writes of static markdown content.
Audit Metadata