Vercel publishes CLI design guidelines for humans and AI agents
JohnPhamous · x · 2026-11-06
Vercel released its CLI Design Guidelines, a manual for building CLIs that people can learn, programs can compose, and agents can operate with confidence. Every command must answer four questions: what it acts on, what will change, what happened (including exit codes), and what's next.
Seven principles include capability-first design (same meaning across CLI/API/SDK/MCP/dashboard), one concept one name, resolve targets before acting (never guess), and keeping interaction, consent, and format independent — prompts need a TTY, --yes grants consent but no data, --json selects format but no consent, and piping changes none of them. Highly relevant now that agents invoke CLIs at scale.
More from coding & agent
- YC boosts stacktrace_ai launch: stop monitoring agents, start steering them in real time — ycombinator · 2026-11-06
- Baseten launches Sandboxes: isolated agent sandboxes that resume state without charging idle time — baseten · 2026-11-06
- Six Claude agents lived 20 days in a simulated town and spontaneously built tools to cooperate — LiminalWanderings · 2026-11-06
- Indie dev builds LLM Cost Guardian to catch per-feature cost spikes in OpenAI/Claude usage — WittyBoysenberry9860 · 2026-11-06
- Drew Breunig's QCon Talk Breaks Down Prompt Debt and Four Context Failures With Concrete Fixes — dbreunig · 2026-11-06
- Dev essay: why a runaway intelligence explosion isn't the risk — 'early onset decoherence' is — round · 2026-11-06