Skills
skills/hmbown/autodialectics/replay/Gen Agent Trust Hub

replay

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the autodialectics CLI to replay runs, which involves executing shell commands populated with user-supplied run_id and policy_id parameters.
  • [EXTERNAL_DOWNLOADS]: The skill documentation instructs the user to install the autodialectics package from a public repository using pip.
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface (Category 8) because it interpolates user-provided identifiers directly into shell commands without explicit sanitization.
  • Ingestion points: run_id and policy_id variables provided by the user.
  • Boundary markers: None identified in the instruction templates.
  • Capability inventory: Local shell command execution via the autodialectics CLI.
  • Sanitization: No validation or escaping of the user-provided IDs is mentioned in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 09:08 AM
Security Audit — agent-trust-hub — replay