skills/xgent-ai/skills/prd/Gen Agent Trust Hub

prd

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted business requirements provided by users and writes them to local files and an external system.
  • Ingestion points: Natural language business requests, ideas, and demands provided by the user (as described in the SKILL.md frontmatter and Step 1).
  • Boundary markers: The skill implements a mandatory "Grilling" discovery protocol (Step 1) and requires explicit user confirmation of a "common understanding" (Step 1.11) before any data is written.
  • Capability inventory: The skill utilizes MCP tools (requirement_create, project_update, testcase_create) to modify an external system (SPMS) and writes markdown PRD documents to the docs/ directory.
  • Sanitization: The agent is instructed to use a "synthesis mode" (Step 4.1) which strictly follows provided templates (references/prd-skeleton.md) and maps data according to a predefined schema (references/spms-mapping.md), explicitly forbidding the invention of new scope.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 03:32 PM
Security Audit — agent-trust-hub — prd