rick-intel

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating a potential surface for indirect prompt injection as part of its primary research function.
  • Ingestion points: Untrusted content is ingested from Azure DevOps work item descriptions and acceptance criteria via 'az boards work-item show', GitHub PR bodies via 'gh pr view', and local markdown files.
  • Boundary markers: The skill uses markdown blockquotes and structured headers in its output to isolate external content.
  • Capability inventory: The agent can execute specific shell commands ('az', 'gh', 'git', 'bd') and perform local file system writes.
  • Sanitization: Instructions include stripping HTML tags from Azure DevOps content before use.
  • [COMMAND_EXECUTION]: The skill relies on external CLI tools to gather data and update system metadata.
  • Evidence: It executes 'az boards work-item show/update', 'gh pr view', 'git branch', and 'bd update'. These tools are used appropriately for searching documentation and linking research to project management items.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 03:04 PM
Security Audit — agent-trust-hub — rick-intel