speckit-specify

Warn

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands, including git fetch, git ls-remote, and a custom bash script located at .specify/scripts/bash/create-new-feature.sh. It also manages the creation and updating of specification and checklist files.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection and shell command injection due to unsafe input interpolation.\n
  • Ingestion points: User-provided descriptions enter via the $ARGUMENTS variable in SKILL.md.\n
  • Boundary markers: Absent; the input is interpolated directly into a bash command string without delimiters.\n
  • Capability inventory: The skill can execute local scripts and perform git operations, providing a path to filesystem and remote repository interaction.\n
  • Sanitization: Relies solely on natural language instructions for the agent to escape single quotes, which is insufficient to prevent command injection via characters like ;, |, or &.\n- [EXTERNAL_DOWNLOADS]: The skill performs git fetch --all --prune, which initiates network connections to external git repositories.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 13, 2026, 06:48 PM