software-architect

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of analyzing untrusted external content.
  • Ingestion points: The skill reads local codebase structures, dependency graphs, and git history in the Phase 1 Context Scan and the Review Process. It also ingests external data from the web via the WebSearch capability.
  • Boundary markers: There are no explicit instructions for the agent to use delimiters or to disregard instructions embedded within the analyzed files or search results.
  • Capability inventory: The skill uses WebSearch for external research, context7 for documentation queries, and the trekker CLI for task and epic management.
  • Sanitization: No sanitization or validation mechanisms are defined for the content extracted from the analyzed projects before it is processed by the agent or used to populate task metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 02:27 AM