Skills
skills/midudev/autoskills/rails-bug-triage/Gen Agent Trust Hub

rails-bug-triage

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of documentation and code skeletons intended for developer assistance. No evidence of malicious intent, credential theft, or unauthorized network activity was found during analysis.
  • [COMMAND_EXECUTION]: The skill refers to executing bundle exec rspec for bug reproduction. This is the primary intended function of the tool and is a standard operation in a Rails development environment.
  • [PROMPT_INJECTION]: The skill processes untrusted bug reports to generate test code, representing a surface for indirect prompt injection. 1. Ingestion points: Bug reports captured in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Execution of rspec shell commands. 4. Sanitization: Absent. This risk is inherent to the TDD process and is mitigated by the assumption of human oversight in the development loop.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 10:26 AM