skills/chrisbanes/grove/using-grove/Gen Agent Trust Hub

using-grove

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes local system commands including git, test, and the grove CLI to manage development workspaces and verify repository state.
  • [EXTERNAL_DOWNLOADS]: Directs users to the author's GitHub repository (github.com/chrisbanes/grove) for the installation of the Grove CLI tool. This is a vendor-owned resource and is considered safe.
  • [REMOTE_CODE_EXECUTION]: Dynamically executes test suites based on detected repository markers (e.g., gradlew, npm, pytest). While this executes code from the local workspace, it is the intended functionality for project validation.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection by parsing JSON output from the grove CLI.
  • Ingestion points: SKILL.md (Step 3: Parsing JSON output for path and id).
  • Boundary markers: Absent.
  • Capability inventory: SKILL.md (Step 4 & 5: Directory navigation and test execution).
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 06:22 PM