diy-harness
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, data exfiltration, or persistence mechanisms were detected. The skill is architected to perform a read-only audit ('Turn A') and provides the user with a task list for review before any actions are taken.
- [EXTERNAL_DOWNLOADS]: The skill documentation references the author's own infrastructure and GitHub repositories for installation instructions and optional companion functionality, which are documented as vendor-owned resources.
- [PROMPT_INJECTION]: The skill ingests project data from local files, creating a surface for indirect prompt injection. 1. Ingestion points: Local repository, CMS, and Carrd project files (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: File writing and project setup operations. 4. Sanitization: Absent. This risk is mitigated by hard-stop instructions that prevent the agent from proceeding to the setup phase without explicit user confirmation of the audit findings.
Audit Metadata