NYC

asc-workflow

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The skill is designed to orchestrate and execute arbitrary shell commands defined under the run key in a .asc/workflow.json file. While this is the intended functionality for automation, it grants the agent the capability to execute system-level commands based on file content.
  • [PROMPT_INJECTION] (LOW): (Category 8: Indirect Prompt Injection)
  • Ingestion points: The skill reads workflow definitions and environment configurations from a repository-local file .asc/workflow.json.
  • Boundary markers: There are no mentioned boundary markers or validation steps to distinguish between legitimate automation commands and malicious instructions embedded in the JSON file.
  • Capability inventory: The tool can execute shell commands, manage environment variables, and chain multiple workflows together.
  • Sanitization: The documentation explicitly mentions that parameters are referenced via shell expansion ($VAR), which is a known risk for command injection if parameters (e.g., BUILD_ID, VERSION) are supplied by an untrusted user or process without proper escaping.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 09:36 PM