cognitive-walkthrough-beads

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface where user input is processed and later used in shell command execution.
  • Ingestion points: User-provided task flows entering through the $ARGUMENTS variable in SKILL.md.
  • Boundary markers: Absent. The skill does not use delimiters to isolate user input from the agent's instructions.
  • Capability inventory: The skill uses the bd CLI (e.g., bd create, bd update) which involves shell command execution.
  • Sanitization: Absent. The agent is instructed to interpolate user input directly into command strings without escaping shell metacharacters, potentially allowing for command injection.
  • [COMMAND_EXECUTION]: The skill leverages the bd command-line utility to perform issue tracking operations and manage project metadata.
  • [EXTERNAL_DOWNLOADS]: The skill documentation suggests installing the beads-cli Python package if the bd command is not available. This is a public tool for task management.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 04:52 PM
Security Audit — agent-trust-hub — cognitive-walkthrough-beads