jewel-release-helper

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes various local build and automation tools including ./gradlew, ./bazel.cmd, and ./tests.cmd. These are standard tools for the specified repository (intellij-community).
  • [COMMAND_EXECUTION]: The skill executes several project-specific Kotlin scripts located in the platform/jewel/scripts/ directory, such as metalava-signatures.main.kts, extract-release-notes.main.kts, and jewel-version-updater.main.kts. These are internal vendor resources used for release automation.
  • [COMMAND_EXECUTION]: The skill performs Git operations including git checkout, git cherry-pick, git tag, and git push. These are necessary for the stated purpose of a release helper and include user confirmation prompts before impactful actions like tagging or pushing.
  • [PROMPT_INJECTION]: The skill ingests untrusted user input via AskUserQuestion (e.g., version numbers, branch names, and issue IDs) and interpolates them into shell commands (e.g., git tag JEWEL-<version>...). While this presents a theoretical command injection surface, the risk is mitigated by the interactive nature of the skill and the context of use by developers within a trusted environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 11:43 AM
Security Audit — agent-trust-hub — jewel-release-helper