skills/asterixcapri/skills/ablate/Gen Agent Trust Hub

ablate

Pass

Audited by Gen Agent Trust Hub on Oct 26, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill suggests that the user install an additional skill from a third-party GitHub repository (mattpocock/skills) using a CLI command (npx skills@latest) if specific design guidance is required. This is a recommendation for the user to act upon and is not an automated or silent execution by the agent.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project source code, which serves as an entry point for untrusted data.
  • Ingestion points: Analyzes project source code and symbols across affected files (SKILL.md).
  • Boundary markers: Employs a 'green signal' requirement where behavior must be verified by tests before and after any changes, and mandates explicit user approval for each modification.
  • Capability inventory: Includes the ability to modify local files and execute shell commands associated with the project's test suite or build pipeline (SKILL.md).
  • Sanitization: Relies on functional verification (tests, linting, type-checking) to ensure that code simplifications do not introduce unexpected behavior, providing a layer of protection against malicious code modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 26, 2026, 09:39 AM