sparc-methodology

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill uses 'npx @claude-flow/cli' across multiple commands. This results in the download and execution of code from the npm registry for an organization that is not on the trusted sources list.
  • COMMAND_EXECUTION (LOW): The scripts 'sparc-init.sh' and 'sparc-review.sh' accept command-line arguments that are used directly in shell commands like 'mkdir', 'touch', and 'ls'. While these scripts have a limited scope, the lack of input sanitization could allow for basic directory traversal or manipulation if malicious paths are provided.
  • PROMPT_INJECTION (LOW): The skill provides an indirect prompt injection surface by interpolating user-provided task descriptions (e.g., [requirements], [feature]) directly into CLI commands. A malicious input could attempt to confuse the underlying agent or the CLI tool, although no explicit jailbreak patterns were detected in the static instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:31 PM