setup-workflow-skills

Fail

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill downloads a shell script from a remote GitHub repository and pipes it directly into the bash interpreter (curl -fsSL ... | bash). This occurs during the setup phase if the required CLI tool is missing.- [COMMAND_EXECUTION]: Executes system commands including bd init for project initialization and sed for modifying the repository's .git/info/exclude file. It also uses command -v to probe the environment for installed binaries.- [EXTERNAL_DOWNLOADS]: Fetches an installation script from an external source (raw.githubusercontent.com/gastownhall).- [PROMPT_INJECTION]: The skill's issue-tracker.md template contains behavioral overrides using markers such as "CRITICAL" and "🚨". These instructions explicitly prohibit the use of standard agent tools (e.g., TodoWrite, TaskCreate) and specific file patterns (MEMORY.md), effectively hijacking the agent's operational workflow in favor of a third-party tool.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/gastownhall/beads/main/scripts/install.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
May 14, 2026, 04:46 PM
Security Audit — agent-trust-hub — setup-workflow-skills