using-jj-workspaces

Pass

Audited by Gen Agent Trust Hub on Feb 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The skill executes shell commands to automate development workflows, including workspace management and environment setup.
  • Evidence: Invokes jj workspace add, npm install, cargo build, and various test runners based on project detection.
  • [EXTERNAL_DOWNLOADS] (LOW): Triggers external network requests to download project dependencies.
  • Evidence: Uses npm, yarn, pnpm, go mod download, and bundle install to fetch packages from public registries.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill is susceptible to indirect injection because it bases execution logic on the contents of untrusted project files.
  • Ingestion points: package.json, Cargo.toml, pyproject.toml, flake.nix, Gemfile, and go.mod.
  • Boundary markers: Absent; the skill blindly trusts the existence of these files to determine which commands to run.
  • Capability inventory: Full command execution (install, build, test) and file modification capabilities.
  • Sanitization: Absent; the skill relies on the underlying package managers to handle potentially malicious configuration files.
  • [DATA_EXPOSURE] (SAFE): Includes a mandatory security check to prevent sensitive workspace data from being tracked by version control.
  • Evidence: Instructions require adding the workspace directory to .gitignore or .jjignore before creating the workspace.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 20, 2026, 06:03 AM