promote-claude-settings

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes project-local configuration data which could contain malicious settings designed to gain global persistence or elevated permissions if promoted by the user.\n
  • Ingestion points: Reads settings from .claude/settings.local.json in the project directory.\n
  • Boundary markers: Implements mandatory interactive confirmation for every proposed change, requiring the user to explicitly approve each setting.\n
  • Capability inventory: Reads and writes configuration files (~/.claude/settings.json) that define agent environment and security boundaries.\n
  • Sanitization: Prompts the user to review and potentially generalize local paths/commands, ensuring the user remains in control of the promoted values.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 03:09 AM