Skills
skills/rhuss/cc-spex/teams-research/Gen Agent Trust Hub

teams-research

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes jq to modify the .claude/settings.local.json file to enable the CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS feature flag. This modification is documented and the user is explicitly prompted to restart the application.
  • [PROMPT_INJECTION]: The skill ingests data from spec.md to identify research topics for sub-agents. This presents an indirect prompt injection surface (Category 8); however, the risk is mitigated as the sub-agents are restricted to read-only operations using tools like Read, Grep, and Glob, and are explicitly forbidden from modifying files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 05:51 PM