Leapmotor embeds ByteDance's TRAE across its 1,000+ engineer R&D org behind a local security perimeter
火山引擎 · wechat · 2026-09-29
Leapmotor partnered with Volcano Engine to embed the AI coding tool TRAE into its full-stack in-house development system. With over 356,000 vehicles delivered globally in H1 2026, R&D scale and complexity have surged — a single smart car involves hundreds of millions of lines of code.
The rollout has three pillars: secure access via enterprise domains and network config so 1,000+ engineers can use AI inside a closed network, with a clear boundary that enterprise code is only used for that inference and never for training; constraining agents via Spec/Plan task decomposition, Rules encoding coding standards and automotive-grade constraints, and Skills codifying team workflows; and efficiency operations through dashboards tracking adoption and model usage, plus per-team model quotas matched to task difficulty and safety levels to control cost.
More from coding & agent
- Generic API-catalog MCP server vs one tool per job: a context-bloat experiment — MandeepKaur9407 · 2026-11-16
- Berkeley paper: LLMs answer with stale values even when the update is in context — ombre-ai · 2026-11-16
- Effective HTML: 4.2k-star agent skills turn wireframes into self-contained interactive prototypes — tom_doerr · 2026-11-16
- skills-gate: one gate skill fronts 148 agent skills at just 389 tokens — Distinct-Pie2389 · 2026-11-16
- Open-weight Karpo-Decision-31B tops Decision Index at 73.9, one forward pass per answer — CodeByPoonam · 2026-11-16
- shadcn boosts @uisdk: one-shot UI rebuild that's instantly E2E-testable by AI — shadcn · 2026-11-16