bootloops-setup

Pass

Audited by Gen Agent Trust Hub on Oct 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to use file system tools to copy directories from the repository to the agent's skill folders. These operations are standard for installation utilities and are restricted to internal repository contents.
  • [PERSISTENCE]: The skill facilitates the installation of agent extensions into standard configuration paths (~/.claude/skills/), which makes them active in subsequent sessions. This persistence is the stated primary purpose of the skill and is gated by explicit user confirmation and scope selection.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user selections from a provided roster to determine which actions to take.
  • Ingestion points: User input during the roster presentation and activation selection in SKILL.md.
  • Boundary markers: Explicit instructions to the agent to never activate without user choice and to confirm all actions.
  • Capability inventory: File system write and directory copying capabilities.
  • Sanitization: The scope of operations is limited to specific pre-existing directories within the local repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 1, 2026, 11:19 PM