create-my-tools-profile
Pass
Audited by Gen Agent Trust Hub on Apr 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs legitimate configuration management by writing rule files to standard directories such as .cursor/rules and .claude/rules. It interacts with the vendor's own Zapier MCP tools and follows user-approved workflows without engaging in suspicious behaviors like unauthorized data transmission or privilege escalation.
- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection by ingesting Zapier tool metadata (names and descriptions) and incorporating it into persistent instruction files. While no specific sanitization or boundary delimiters are used to isolate this untrusted content, the risk is considered safe within the context of the skill's primary purpose and the user's control over the data source. Ingestion points: list_enabled_zapier_actions; Boundary markers: None; Capability inventory: File system writes; Sanitization: None.
Audit Metadata