triage

Pass

Audited by Gen Agent Trust Hub on Sep 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data from SIT or UAT results (CSV, Markdown table, or spreadsheet) to classify items and generate GitHub issues. Maliciously crafted data in these source files could attempt to influence the agent's classification logic or content generation for the issues.
  • Ingestion points: Step 1 in SKILL.md identifies various input formats including CSV, Markdown, and XLSX files.
  • Boundary markers: The instructions do not define explicit delimiters to isolate untrusted data, though they do instruct the agent to map specific columns.
  • Capability inventory: The skill can create or comment on GitHub issues (via gh CLI), manage labels, and execute a local Python script.
  • Sanitization: The included script scripts/check_triage.py validates the resulting log file for ASCII-only content and structural consistency, which serves as a post-generation integrity check.
  • [COMMAND_EXECUTION]: The skill executes a local Python script scripts/check_triage.py using python3 to validate logs. It also utilizes the GitHub CLI (gh) for listing and managing repository issues and labels.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 30, 2026, 02:49 AM