skill-builder

Pass

Audited by Gen Agent Trust Hub on Nov 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted user input (workflows, steps, or existing skill files) and transform it into executable agent instructions. This workflow presents an attack surface where malicious instructions embedded in the input could influence the generated output or the agent's behavior during the creation process.
  • Ingestion points: Processes user-provided text describing workflows, operating procedures, and existing SKILL.md files as specified in the 'Contract' section of the SKILL.md file.
  • Boundary markers: The instructions do not mandate the use of delimiters or 'ignore embedded instructions' warnings when handling the untrusted input text during the generation phase.
  • Capability inventory: The skill's stated purpose involves file-writing capabilities to create or overwrite SKILL.md files in the local filesystem.
  • Sanitization: There are no explicit requirements for sanitizing, escaping, or validating the external content before it is incorporated into the generated skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 18, 2026, 01:28 AM