skills/davekilleen/dex/project-health/Gen Agent Trust Hub

project-health

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the ls -la command to list contents of the 04-Projects/ directory to identify active project files and folders.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes user-controlled content from project notes and weekly plans. Ingestion points: Project files within the 04-Projects/ folder and the 00-Inbox/Weekly_Plans.md file. Boundary markers: None identified; project content is scanned directly for keyword triggers without delimiters or instructions to ignore embedded commands. Capability inventory: The skill performs directory listings, file reads, local file modifications to System/usage_log.md, and metadata telemetry via the track_event tool. Sanitization: No sanitization or content validation is performed on the processed project data.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 12:41 AM
Security Audit — agent-trust-hub — project-health