skills/rheinmir/skills/fdk-uat/Gen Agent Trust Hub

fdk-uat

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill fetches and executes a bootstrap script directly from GitHub's raw content service (raw.githubusercontent.com). This is used in the UAT process to verify that the remote installation path works correctly for new framework versions.
  • [COMMAND_EXECUTION]: The skill executes shell commands to perform repository maintenance tasks, including creating and deleting remote Git branches (git push origin --delete) and performing forced resets or reverts (git push --force-with-lease) to restore the repository state if a smoke test fails.
  • [EXTERNAL_DOWNLOADS]: Downloads configuration files and bootstrap scripts from raw.githubusercontent.com during the UAT lifecycle.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill ingests and processes content from remote repositories, including the bootstrap.sh script and policy.yaml rules, which are executed or parsed during testing.
  • Boundary markers: No explicit delimiters or 'ignore embedded instructions' warnings are applied to the external content being processed.
  • Capability inventory: The skill possesses extensive shell access, including the ability to run curl, git, python3, and the orca CLI, alongside local filesystem and remote Git state modification privileges.
  • Sanitization: No sanitization or integrity validation of the downloaded script content is performed before execution via the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 02:26 AM