voter-calibration
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill safely delegates its processing entirely to a local Rust binary wrapper script (
${CLAUDE_PLUGIN_ROOT}/scripts/larch.sh). It does not invoke any unvalidated inputs, download remote untrusted payloads, or perform unsafe dynamic command evaluations. The data ingestion handles repository logs offline or structured metadata query paths via an established tool pipeline.
Audit Metadata