scribe
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured templates and workflows for creating PRDs, SRSs, and other technical documents. It incorporates security-relevant sections such as Non-Functional Requirements (NFRs) for security (e.g., authentication, audit logging) and operational runbook guidelines.
- [EXTERNAL_DOWNLOADS]: The skill references several well-known third-party tools and services for format conversion and API documentation, including Pandoc, LibreOffice, WeasyPrint, Typst, Speakeasy, and Mintlify. These references are consistent with the skill's purpose of document processing and publication.
- [COMMAND_EXECUTION]: The skill provides template shell commands for use with documentation tools (e.g.,
pandoc,soffice,weasyprint). These are standard operational instructions for the agent to perform its intended tasks and do not involve arbitrary or malicious command execution. - [INDIRECT_PROMPT_INJECTION]: The skill includes instructions for authoring AI-agent-consumable specifications (AGENTS.md convention). While it processes external data such as user requirements, it follows structured elaboration workflows and quality gates that serve as boundary markers, minimizing the risk of accidental instruction obedience from untrusted data.
Audit Metadata