brief-to-tasks

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests untrusted external design brief documents and project files.
  • Ingestion points: Reads file contents from .design/*/DESIGN_BRIEF.md, INFORMATION_ARCHITECTURE.md, and package.json in the local workspace.
  • Boundary markers: No specific boundary delimiters or isolation guards are established to prevent malicious instructions embedded in a design brief from manipulating the output generation.
  • Capability inventory: The skill's capabilities are limited to reading project files and writing a markdown file (TASKS.md) into the project directory. It lacks high-risk capabilities such as network communication, command execution, or dynamic code evaluation.
  • Sanitization: No sanitization or validation of the input file text is specified before it is integrated into the prompt context for generating tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:40 PM
Security Audit — agent-trust-hub — brief-to-tasks