adoption-plan

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a set of documentation-focused instructions that guide an agent through a design system adoption process. It operates on local files and uses provided tools to query established design system catalogs.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from the product codebase and issue trackers, creating a surface for indirect prompt injection. However, the risk is low as the output is descriptive and the agent is instructed to verify all findings against a central catalog.
  • Ingestion points: Reads local source code, templates, CSS, and issue tracker lists (via ADOPTION-INTAKE.md and phase-1-teardown.md).
  • Boundary markers: Employs provenance tagging ([verified] vs [reported]) to distinguish evidence from raw input (SKILL.md).
  • Capability inventory: Reads repository files, writes analysis to local Markdown files, and invokes command-line tools like gh.
  • Sanitization: No technical sanitization of ingested code comments or issue text is mentioned, but findings are subject to manual review via the produced reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 01:34 PM
Security Audit — agent-trust-hub — adoption-plan