jujutsu-colocated

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • Data Exposure & Exfiltration (SAFE): The skill explicitly forbids accessing or editing .env or environment variable files, mitigating risks of accidental credential exposure.
  • Indirect Prompt Injection (SAFE): The skill processes untrusted data from VCS repositories.
  • Ingestion points: Files like SKILL.md and references/colocated-workflows.md include commands like jj status, jj diff, and jj log that read repository content.
  • Boundary markers: None explicitly defined in the provided command strings.
  • Capability inventory: VCS management including pushing to remotes and workspace creation.
  • Sanitization: Relies on the agent following behavioral safety rules rather than automated input sanitization. This is considered SAFE as it is the primary intended purpose of a VCS skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:27 PM