- type
- Coding agent
- interface
- CLI
- openSource
- Yes (MIT)
- lineage
- Plan-driven multi-agent coding loop (`neal run/plan/execute`) with separate planner, coder, and reviewer roles, each on a provider/model the user picks; coder and reviewer are deliberately run on different vendors so review is adversarial rather than one model grading its own work; each scope starts the coder from a fresh context to prevent drift across long tasks; reviewer access is enforced read-only across provider adapters
- note
- Installs via npm (@navels/neal). Distinct from other multi-agent CLIs (Deep Agents Code, CodeWhale, Huawei CodeArts Agent Team) by its explicit cross-vendor adversarial-review primitive rather than same-vendor sub-agents; GitHub repo created 2026-07-27, very early-stage single-maintainer project
- iconUrl
- /research/harness-icons/neal.png
