design-prd
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a standard documentation workflow including project auto-detection, competitive research, and PRD generation. All automated actions (globbing files, reading package manifests, querying database schemas) are scoped to the local project environment for the purpose of requirement gathering.
- [EXTERNAL_DOWNLOADS]: The skill uses
firecrawl_searchandfirecrawl_scrapefor competitive research andcontext7for technical feasibility. These are purpose-built tools for the agent platform to interact with the web and documentation, and their use here is aligned with the skill's primary function of product research. - [COMMAND_EXECUTION]: The skill instructs the agent to run standard read-only commands for stack detection (e.g., reading
package.json, globbing route directories, and inspecting SQL migrations). These are routine operations for an AI agent to understand its environment. - [DATA_EXPOSURE]: While the skill accesses database schemas (Supabase/Prisma/SQL), it does so to understand the data model for the PRD, not to extract sensitive user data. No hardcoded credentials or exfiltration patterns were detected.
Audit Metadata