lead-researcher

SUSPICIOUS: the skill’s stated purpose matches a research-orchestration role, but it expands trust by explicitly delegating to many other skills and may process untrusted external research content while potentially triggering replication/execution through sub-skills. No direct credential harvesting, exfiltration endpoint, or mismatched data flow is present here; the main risks are transitive skill trust, prompt-injection exposure from external content, and minor official-installer supply-chain risk around optional Ollama use.