team-combat

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill follows a structured orchestration pattern using the Task tool to delegate specialized work to subagents. All file operations are restricted to project-specific paths for configuration and documentation. No evidence of obfuscation, remote code execution, or unauthorized data exfiltration was found.- [PROMPT_INJECTION]: The skill ingests user-supplied combat feature descriptions and reads from local project files like production/review-mode.txt and .claude/docs/technical-preferences.md to determine workflow behavior and subagent types. This constitutes a surface for indirect prompt injection.
  • Ingestion points: User-provided feature description argument, production/review-mode.txt, and .claude/docs/technical-preferences.md.
  • Boundary markers: None present.
  • Capability inventory: Orchestrates subagents via the Task tool and has access to Bash, Read, Write, and Edit tools.
  • Sanitization: None identified; the skill passes raw configuration and description data directly to subagents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 04:19 PM
Security Audit — agent-trust-hub — team-combat