the-workflow-builder

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No instructions were found attempting to bypass safety filters, override system behavior, or extract system prompts. The skill uses benign instructional language to guide the workflow design process.\n- [DATA_EXPOSURE]: Access to local files is restricted to .agents/product-context.md, which is a standard pattern for providing project context to AI agents. No sensitive credentials, private keys, or system-level configuration paths are accessed.\n- [REMOTE_CODE_EXECUTION]: The skill does not contain any patterns for downloading external scripts, installing unverified packages, or executing code at runtime. It functions purely as a text-based specification generator.\n- [OBFUSCATION]: No hidden content, Base64 encoding, zero-width characters, or homoglyph substitutions were detected. The instructions and reference files are fully transparent and human-readable.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data and context files to build specifications. However, since the skill only produces a text document (a workflow spec) and lacks tools for automated file modification or command execution, there is no exploitable capability chain present.\n- [DYNAMIC_CONTEXT_INJECTION]: No use of shell command interpolation (!command) was found in the skill definitions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 12:03 PM