maa-workflow-build

Pass

Audited by Gen Agent Trust Hub on Nov 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user intentions to drive the orchestration of automation workflows and the generation of pipeline nodes.\n
  • Ingestion points: Ingests user-provided intent in the SPECIFY phase (SKILL.md).\n
  • Boundary markers: Lacks explicit delimiters for user input, relying instead on the transformation of intent into a structured YAML task contract.\n
  • Capability inventory: Capable of generating executable MaaFramework pipeline configurations and executing shell commands via the $maa-cli-operate tool.\n
  • Sanitization: No specific text-based sanitization or filtering of user input is documented prior to its use in contract compilation.\n- [COMMAND_EXECUTION]: The skill integrates with command-line tools to perform environmental checks and runtime operations.\n
  • Evidence: References the use of $maa-cli-operate for repeatable checks or guarded runtime operations during the implementation and verification phases.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 22, 2026, 01:28 AM