<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs" -->

---
title: Giving AI agents design system context from the...
description: Atlassian describes why it added a command-line interface alongside its existing MCP server and agent skill for delivering design system context to AI coding...
canonical: https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: Giving AI agents design system context from the terminal: what we learned building a CLI | daily.dev
og:description: Atlassian describes why it added a command-line interface alongside its existing MCP server and agent skill for delivering design system context to AI coding...
og:url: https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs
og:image: https://api.daily.dev/og/posts/OZVvf6crS.png
og:image:alt: Giving AI agents design system context from the terminal: what we learned building a CLI
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Giving AI agents design system context from the terminal: what we learned building a CLI

**[Atlassian](https://daily.dev/sources/atlassian)** · 8 min read · 0 upvotes · 0 comments

## Summary

Atlassian describes why it added a command-line interface alongside its existing MCP server and agent skill for delivering design system context to AI coding agents. The CLI reaches agents lacking an MCP client, and benchmarking against real front-end tasks showed sessions completed 8% faster (352s to 325s), token usage dropped 8% (218k to 201k), and CLI adoption surpassed MCP usage by 60%. Key fixes included moving off npx to an internal pre-installed plugin (cutting startup overhead by 95%), adding batch queries, and guiding agents toward follow-up commands. The team now defaults to CLI, falling back to MCP only when terminal access is unavailable.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.atlassian.com/blog/ai-at-work/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli>

## Questions this post answers

### Why would a team build a CLI for AI agents when they already have an MCP server?

A CLI reaches agents and environments where an MCP client isn't configured, such as local development, scripts, CI, and coding agents without an MCP setup. Atlassian's benchmarking found their CLI also completed tasks 8% faster (352s vs 325s) and used 8% fewer tokens (218k vs 201k) than their MCP server baseline, with agents adopting it 60% more than MCP.

_daily.dev surfaces engineering write-ups like this for teams weighing MCP versus CLI for agent tooling._

### How much can moving from npx to a pre-installed CLI plugin reduce agent startup overhead?

Switching from npx, which downloaded and booted a package on every call with over 15 seconds of overhead per query in a monorepo, to a pre-installed plugin in an internal CLI reduced that overhead to milliseconds, a roughly 95% reduction. npx is still kept as a fallback where the internal CLI isn't available.

_teams optimizing agent tool latency can find similar performance case studies on daily.dev._

## Similar posts on daily.dev

- [Agentic Terminal - How Your Terminal Comes Alive with CLI Agents](https://daily.dev/posts/agentic-terminal---how-your-terminal-comes-alive-with-cli-agents-ygueyoqlj) · InfoQ · 1 upvotes · 0 comments
- [MCP vs. CLI: Which Is Better for Agentic AI?](https://daily.dev/posts/mcp-vs-cli-which-is-better-for-agentic-ai--rwtvgc6z0) · Nordic APIs · 0 upvotes · 0 comments
- [Teaching AI to speak our design language](https://daily.dev/posts/teaching-ai-to-speak-our-design-language-qcj2tdmqh) · Atlassian · 0 upvotes · 0 comments
- [“The terminal still matters”: Amp rebuilds its CLI for an agentic future beyond the command line](https://daily.dev/posts/the-terminal-still-matters-amp-rebuilds-its-cli-for-an-agentic-future-beyond-the-command-line-zkz0cwa0g) · The New Stack · 1 upvotes · 0 comments
- [What I learned building an opinionated and minimal coding agent](https://daily.dev/posts/what-i-learned-building-an-opinionated-and-minimal-coding-agent-myqj9tthj) · Hacker News · 1 upvotes · 0 comments

---

Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#cli](https://daily.dev/tags/cli), [#mcp](https://daily.dev/tags/mcp), [#design-systems](https://daily.dev/tags/design-systems), [#atlassian](https://daily.dev/tags/atlassian)

[View this post on daily.dev](https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"Giving AI agents design system context from the terminal: what we learned building a CLI","url":"https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs"},"datePublished":"2026-09-16T22:37:20.313Z","dateModified":"2026-09-18T18:09:01.582Z","description":"Atlassian describes why it added a command-line interface alongside its existing MCP server and agent skill for delivering design system context to AI coding...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/3c662a3fd505c7ace88f878e3bccee5c?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/3c662a3fd505c7ace88f878e3bccee5c?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Atlassian","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Atlassian","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/44750d8e66de4411bb477159b321bb43","url":"https://daily.dev/sources/atlassian"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,cli,mcp,design-systems,atlassian","timeRequired":"PT8M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Atlassian","item":"https://daily.dev/sources/atlassian"},{"@type":"ListItem","position":3,"name":"Giving AI agents design system context from the terminal: what we learned building a CLI"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/giving-ai-agents-design-system-context-from-the-terminal-what-we-learned-building-a-cli-ozvvf6crs#faq","mainEntity":[{"@type":"Question","name":"Why would a team build a CLI for AI agents when they already have an MCP server?","acceptedAnswer":{"@type":"Answer","text":"A CLI reaches agents and environments where an MCP client isn't configured, such as local development, scripts, CI, and coding agents without an MCP setup. Atlassian's benchmarking found their CLI also completed tasks 8% faster (352s vs 325s) and used 8% fewer tokens (218k vs 201k) than their MCP server baseline, with agents adopting it 60% more than MCP. daily.dev surfaces engineering write-ups like this for teams weighing MCP versus CLI for agent tooling."}},{"@type":"Question","name":"How much can moving from npx to a pre-installed CLI plugin reduce agent startup overhead?","acceptedAnswer":{"@type":"Answer","text":"Switching from npx, which downloaded and booted a package on every call with over 15 seconds of overhead per query in a monorepo, to a pre-installed plugin in an internal CLI reduced that overhead to milliseconds, a roughly 95% reduction. npx is still kept as a fallback where the internal CLI isn't available. teams optimizing agent tool latency can find similar performance case studies on daily.dev."}}]}
```

