faion-software-architect

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill is designed to automatically investigate local project files (e.g., ADRs, Docker Compose, and Kubernetes manifests) to determine the architecture of a project. This pattern creates a surface for indirect prompt injection.
  • Ingestion points: As defined in SKILL.md, the agent uses Glob and Grep to read content from various configuration files (docker-compose*.yml, k8s/**/*.yaml, etc.) and documents (adr/*.md).
  • Boundary markers: The skill does not provide the agent with specific delimiters or instructions to treat data from these project files as untrusted or to ignore embedded directives.
  • Capability inventory: The agent possesses Write, Edit, and WebFetch tools, which could be leveraged if an injection in a project file successfully influences the agent's behavior.
  • Sanitization: There is no evidence of sanitization, validation, or escaping of the content read from the local filesystem during the discovery phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:15 PM